Purpose

The ABILHAND is a semi-structured item-response questionnaire that measures manual ability according to an individual’s perceived difficulty performing daily bimanual tasks.

In-Depth Review

Purpose of the measure

The ABILHAND is an interview-based assessment tool that measures a patient’s perceived difficulty using his/her hands to perform manual activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
in daily life. The ABILHAND assesses active function of the upper limbs. The tool measures an individual’s ability to perform bimanual tasks, regardless of strategies used to complete the task (Ashford et al., 2008; Penta et al., 1998)

Available versions

The ABILHAND was originally developed by Penta et al. (1998) as a 56-item, 4-level questionnaire of unimanual and bimanual ability for patients with rheumatoid arthritis. The original ABILHAND was intended to measure rehabilitation outcomes and to provide guidelines for goal setting in treatment planningPlanning ability involves anticipating future events, formulating a goal or endpoint, and devising a sequence of steps or actions that will achieve the goal or endpoint" (Anderson, 2008, p. 17)
(Gustafsson et al., 2004). Penta et al. (2001) found that patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. were able to complete unimanual activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
with the unaffected limb, regardless of hand dominance, whereas bimanual tasks were more difficult. Accordingly, a version was developed specifically for patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. that only included bimanual items, as well as two alternate unimanual’ activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
that require skillful use of the affected hand (cutting nails, filing nails). Penta et al. (2001) also reviewed the 4-level scoring criterion (impossible, very difficult, difficult, easy) and found that patients rarely used the very difficult’ score. This indicated that the two intermediate scoring criteria (very difficult, difficult) were not sufficiently differentially distinct. Accordingly, the strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. version of the ABILHAND was developed with a 3-level scoring criterion (impossible, any difficulty, easy).

Other impairment-specific versions were subsequently created with modified item sets and levels. Each version of the ABILHAND has its own Rasch-derived item difficulty calibrations that rely on computerized algorithms to obtain the patient’s overall measure from his/her responses (Simone et al., 2011).

Features of the measure

Items:

The ABILHAND is an inventory of 23 bimanual activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
(from most difficulty to least difficult):

1. Hammering a nail
2. Threading a needle
3. Peeling potatoes with a knife
4. Cutting own nails
5. Wrapping up gifts
6. Filing own nails
7. Cutting meat
8. Peeling onions
9. Shelling hazel nuts
10. Opening a screw-topped jar
11. Fastening zipper of jacket
12. Tearing open pack of chips
13. Buttoning up a shirt
14. Sharpening a pencil
15. Spreading butter on a slice of bread
16. Fastening a snap
17. Buttoning up trousers
18. Taking the cap off a bottle
19. Opening mail
20. Squeezing toothpaste on a toothbrush
21. Pulling up the zipper of trousers
22. Unwrapping a chocolate bar
23. Washing hands

Scoring:

The patient is asked to rate his/her perceived difficulty performing items without help, according to the following scoring criteria:

• 0 = impossible
• 1 = difficult
• 2 = easy

Tasks that the patient has not performed in the past 3 months are not scored and are encoded as missing responses.

The ABILHAND was developed using the Rasch measurement model, which provides a method to convert the ordinal raw score into a linear measure on a unidimensional scale. Item scores are entered into the WINSTEPS computer program, and raw ordinal data is converted to linear measures expressed in logits (log-odds probability units). The total score is scaled along a unidimensional continuum with 0 at the centre of the scale, whereby the higher the logit number, the greater the patient’s perceived ability (Gustafsson et al., 2004).

What to consider before beginning:

Users should note that self-estimated measures (i.e. when scores are not based on clinician observation of performance) are subject to overestimation or underestimation of actual performance, depending on motivation and cognitive skills (Penta et al. 2001).

Clinicians should consider patient factors such as self-esteem, insight, vision, hearing, language and cognitive function prior to administering the ABILHAND (Gustafsson et al., 2004).

Mpofu & Oakland (2010) advise caution when using the ABILHAND to measure improvements in impairment of the affected upper limb after strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. rehabilitation. The ABILHAND does not take into consideration the arm used to perform a task or compensatory strategies employed to complete the task. Accordingly, improvement in scores may be based on use of compensatory strategies rather than on improvement in the affected arm.

Time:

The ABILHAND takes 10 to 30 minutes to administer (Ashford et al., 2008; Connell et al., 2012).

Training requirements:

No training requirements have been specified for the ABILHAND, although administration by a clinician is recommended (Ashford et al., 2008).

Equipment:

The ABILHAND is a semi-structured questionnaire that does not require specific equipment, however the WINSTEPS computer program is required to process raw scores.

Client suitability

Can be used with:

• Individuals with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.
• Individuals with rheumatoid arthritis
• Individuals with systemic sclerosis

Should not be used with:

• Due to the subjective nature of the patient’s reports, this measure should not be used with individuals with severe cognitive deficits (Penta et al., 2001).
• The ABILHAND may not be suitable for use with patients with aphasiaAphasia is an acquired disorder caused by an injury to the brain and affects a person’s ability to communicate. It is most often the result of stroke or head injury.
  An individual with aphasia may experience difficulty expressing themselves when speaking, difficulty understanding the speech of others, and difficulty reading and writing. Sadly, aphasia can mask a person’s intelligence and ability to communicate feelings, thoughts and emotions. (The Aphasia Institute, Canada) or apraxia (Gustafsson et al., 2004).

In what languages is the measure available?

• French
• English
• Dutch
• Italian
• Swedish

Summary

What does the tool measure?	Manual ability of the upper extremity.
What types of clients can the tool be used for?	The ABILHAND can be used with, but is not limited to, patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
Time to administer	10-30 minutes
Versions	AH-RA for rheumatoid arthritis (46 items, 4 levels) AH-RA revised version (27 items, 3 levels) ABILHAND-ULA for upper limb amputees (22 items; 4 levels) SSC-adapted ABILHAND for systemic sclerosis (26 items, 3 levels) ABILHAND – neuromuscular age-independent version (22 items) ABILHAND-Kids (21 items)
Other Languages	French, English, Swedish, Dutch, Italian
Measurement Properties
ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: – Order of difficulty of items has been confirmed by Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model. . – One study reported a high item reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest . index. – One study reported high person separation reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest . . Test-retest: No studies have reported on the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of the ABILHAND. Intra-rater: No studies have reported on the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the ABILHAND. Inter-rater: No studies have reported on the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the ABILHAND.
ValidityThe degree to which an assessment measures what it is supposed to measure.	Content: – One study reported that the 23 items of the ABILHAND define a common continuum of manual ability, and items are coherent with the overall questionnaire and contribute to the measurement of manual ability. – One study examined stability of item difficulty of the ABILHAND and found that item hierarchy was substantially retained across different groupings (impairment, age, sex, ability). – One study reported that scores explained 84% of observed variance. The main factor across the residuals explained only 11.4% of the residual variance (1.8% of the total variance). Criterion: Concurrent: One study examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.” of the ABILHAND among patients with chronic upper limb impairment resulting from conditions including strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and reported adequate correlations with the Box and Block Test, Jamar handgrip and Purdue pegboard test, and an adequate negative correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with the Nine Hole Peg Test. Predictive: No studies have reported on the predictive validityA form of criterion validity that examines a measure’s ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. of the ABILHAND. Construct: Convergent/Discriminant: No studies have reported on the convergent/discriminant validityThe degree to which an assessment measures what it is supposed to measure. of the ABILHAND. Known Groups: – One study reported highly significant differences in ABILHAND scores between patients with tetraparesis, hemiparesis, other neurological impairments (multiple sclerosis, Parkinson’s disease, ataxia) and healthy subjects. – One study reported no correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between ABILHAND scores and country, age, sex, time since strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., affected side, lesion site or tactile sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” ; poor correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with grip strength and manual dexterity of the unaffected limb; poor negative correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely “pull themselves together” and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression. ; adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with grip strength and manual dexterity of the affected limb; and excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with upper limb motricity.
Floor/Ceiling Effects	No studies have reported on the floor/ceiling effects of the ABILHAND.
Does the tool detect change in patients?	– No studies have reported on the responsivenessThe ability of an instrument to detect clinically important change over time. of the ABILHAND. – One study reported that the ABILHAND demonstrates 92% sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” and 80% specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative). at a lower cutoff score of 80/100.
Acceptability	The ABILHAND is non-invasive and quick to administer. The items are considered reflective of real-life activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function. (i.e. ecologically valid).
Feasibility	The ABILHAND is portable and is suitable for administration in various settings. The assessment is quick to administer and requires minimal specialist equipment or training.
How to obtain the tool?	The ABILHAND is available in Penta, M., Tesio, L., Arnould, C., Zancan, A., & Thonnard, J-L. (2001). The ABILHAND questionnaire as a measure of manual ability in chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. patients: Rasch-based validation and relationship to upper limb impairment. StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., 32, 1627-34

Psychometric Properties

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the ABILHAND. While additional studies have been conducted on other ABILHAND versions, this review specifically addresses the psychometric properties of the 23-item strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. version of the ABILHAND, unless otherwise specified. Two studies were identified.

Floor/Ceiling Effects

No studies have reported on the floor or ceiling effects of the ABILHAND. However, given the hierarchical relationship of items, lower-level tasks of the ABILHAND may be susceptible to floor effects (Ashford et al., 2008).

Reliability

Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.:
Penta et al. (2001) examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the original 56-item ABILHAND in a sample of 103 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
and reported high reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
(Rasch separation reliability=0.90; person separation reliability=0.90). The authors examined the stability of the scale through differential item functioning (DIF) tests among 12 subgroups: sex (male/female); country (Belgium/Italy); age (< 60/≥ 60), affected side (dominant/nondominant); delay since strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (< 2 years/≥ 2 years), level of depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely “pull themselves together” and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
, dexterity and manual ability of the unaffected limb, grip strength, dexterity and sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
of the affected limb, and motricity of the affected limb. The difficulty hierarchy of the ABILHAND was uniformly perceived by patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Simone et al. (2011) examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the ABILHAND in a sample of 126 patients with chronic upper limb impairment resulting from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=83), multiple sclerosis (n=17), peripheral or cerebellar ataxia (n=13), spinal cord lesion (n=10) or Parkinson’s disease (n=3), and 24 health subjects. The ABILHAND demonstrated high reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
(item reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
index=0.94; Cronbach’s alpha=0.99). All items of the ABILHAND fit the Rasch model satisfactorily. There were at least 4 strata of statistically different measures, indicating that variance across scores did not reflect randomness. The authors also examined stability of item difficulty through differential item functioning (DIF) by comparing 4 different groupings of the sample pool: impairment (hemiparesis vs. other); age (≤ 69 vs. > 69); sex (male vs. female); and ability (above median vs. below median). There was a very moderate DIF across the grouping criteria, whereby item hierarchy was substantially retained for all subgroups: impairment (1 outlier: buttoning a shirt); sex (6 outliers: fastening a snap, shelling hazel nuts, hammering a nail, wrapping up gifts, peeling potatoes, spreading butter); age (4 outliers: threading a needle, wrapping up gifts, spreading butter, fastening a snap); and ability (2 outliers: sharpening a pencil, cutting meat).

Test-retest:
No studies have reported on the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the ABILHAND.

Intra-rater:
No studies have reported on the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
of the ABILHAND.

Inter-rater:
No studies have reported on the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the ABILHAND. Note, however that inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
is less necessary because administration of the ABILHAND does not rely on clinician-observation of patient performance.

Validity

Content:

Penta et al. (2001) examined the measure of perceived difficulty of the ABILHAND in a sample of 103 patients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Item distribution ranged from 1.72 to -2.18 logits. All items fit the Rasch model and the 23 items define a common continuum of manual ability. All point measure correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients (RPM) were positive, indicating that all items are coherent with the overall questionnaire and contribute to the measurement of manual ability. Although fit statistics indicated that most activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
adequately measure recovery of manual ability in chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., 1 item obtained an outlier outfit value (buttoning up a shirt, mean square=1.64), and four items obtained outlier infit values (cutting meat, mnsq=0.69; shelling hazel nuts, mnsq=1.33; tearing open a packet of chips, mnsq=1.22; sharpening a pencil, mnsq=0.65).

Penta et al. (2001) examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension.
of the ABILHAND by comparing the ranking of item difficulty with expert opinion of four occupational therapists regarding the involvement of the affected hand in each activity. The following classifications were used: (1) the item does not require the affected limb, if it is broken down into several unimanual sequences; (2) the task requires the affected upper limb to stabilize an object but does not involve any fingers; and (3) the task requires precision grip, grip strength, dexterity or any digital activity from the affected side. Findings indicate that more difficult items also tend to require a greater degree of use of the affected limb, whereas easier items do not require the use of the affected limb.

Simone et al. (2011) examined the validityThe degree to which an assessment measures what it is supposed to measure.
of the ABILHAND in a sample of 126 patients with chronic upper limb impairment resulting from strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. (n=83), multiple sclerosis (n=17), peripheral or cerebellar ataxia (n=13), spinal cord lesion (n=10) or Parkinson’s disease (n=3), and 24 health subjects. Modeled scores explained 84% of observed variance. The main factor across the residuals explained only 11.4% of the residual variance (1.8% of the total variance).

Criterion:

Concurrent:
Simone et al. (2011) compared the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the ABILHAND, Jamar handgrip, Box and Block Test (BBT), Purdue pegboard test and Nine Hole Peg Test (NHPT) in a sample of 126 patients with chronic upper limb impairment resulting from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, sensory or cerebellar ataxia, spinal cord lesion or Parkinson’s disease, and 24 healthy subjects, using Pearson’s r. Adequate correlations were found between the ABILHAND and the Jamar handgrip (r=0.377, p=0.001), BBT (r=0.481, p=0.000) and the Purdue pegboard test (r=0.493, p=0.000), and an adequate negative correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
was found between the ABILHAND and the NHPT (r=-0.370, r=0.007).

Predictive:
No studies have reported on the predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the ABILHAND.

Construct:

Convergent/Discriminant:
No studies have reported on the convergent/discriminant validityThe degree to which an assessment measures what it is supposed to measure.
of the ABILHAND.

Known Group:
Penta et al. (2001) examined the relationship of the ABILHAND measures to other demographic and clinical variables in a sample of 103 patients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., using univariate ANOVA and correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients (Mann-Whitney U test, Kruskal-Wallis H tests, Spearman p, Pearson r). Tests revealed no significant differences in ABILHAND measures according to demographic indexes of country (Belgium/Italy), sex or age. Clinical variables such as time since strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., affected side (dominant/nondominant), lesion site and tactile sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also "Specificity."
of either limb (measured using the Semmes-Weinstein tactile sensation test) were not significantly related to ABILHAND measures. There was a poor correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between ABILHAND measures and grip strength (Jamar handgrip, R=0.242, P<0.014) and manual dexterity (Box and Block Test, R=0.248, P=0.012) of the unaffected limb, and a poor negative correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
(Geriatric DepressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
Scale, p=-0.213, P=0.030). ABILHAND measures demonstrated an adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with grip strength (R=0.562, P<0.001) and manual dexterity (R=0.598, P<0.001) of the affected limb, and an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with upper limb motricity (Brunnstrom upper limb motricity test, p=0.730, P<0.001). Results showed a direct relationship between ABILHAND measures of manual ability and impairment on the affected side, where more complex combinations of manual dexterity without/without grip strength and/or upper limb motricity impairment correlated with higher manual disability.

Simone et al. (2011) examined the known-group validityThe degree to which an assessment measures what it is supposed to measure.
of the ABILHAND in a sample of 126 patients with chronic upper limb impairment resulting from strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, sensory or cerebellar ataxia, spinal cord lesion or Parkinson’s disease, and 24 healthy subjects, using Kruskal-Wallis test. Highly significant differences (P<0.001) were found between patients with tetraparesis, hemiparesis, other neurological impairments (multiple sclerosis, Parkinson’s disease, ataxia) and control participants.

Responsiveness

Simone et al. (2011) reported a satisfactory match between the distribution of item difficulty levels and patients’ ability levels. The average ability of healthy controls vs. patients with chronic upper limb impairment resulting from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, sensory or cerebellar ataxia, spinal cord lesion or Parkinson’s disease was 89 (standard error=8) vs. 63 (standard error=17).

SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
& SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
:
Simone et al. (2011) examined the sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
and specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
of the ABILHAND in a sample of 126 patients with chronic upper limb impairment resulting from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, sensory or cerebellar ataxia, spinal cord lesion or Parkinson’s disease, and 24 healthy subjects. An “impairment-normality” cut-off was computed through logistic regression and a lower cut-off score of 80/100 is proposed for healthy controls (area under ROC curve=0.9097, p<0.05). This allowed correct classification of patients vs. healthy controls with a 92% sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
rate and 80% specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
rate, whereby 82% of the sample was correctly classified.

References

• Ashford, S., Slade, M., Malaprade, F., & Turner-Stokes, L. (2008). Evaluation of functional outcome measures for the hemiparetic upper limb: a systematic review. Journal of Rehabilitation Medicine, 40, 787-95.
• Connell, L.A. & Tyson, S.F. (2012). Clinical reality of measuring upper-limb ability in neurological conditions: a systematic review. Archives of Physical Medicine and Rehabilitation, 93, 221-8.
• Gustafsson, S., Sunnerhagen, K.S, & Dahlin-Ivanoff, D. (2004). Occupational therapists’ and patients’ perceptions of ABILHAND, a new assessment tool for measuring manual ability. Scandinavian Journal of Occupational Therapy, 11, 107-17.
• Mpofu, E. & Oakland, T. (2010). Rehabilitation and Health Assessment: Applying ICF Guidelines. New York: Springer Publishing Company.
• Penta, M., Tesio, L., Arnould, C., Zancan, A., & Thonnard, J-L. (2001). The ABILHAND questionnaire as a measure of manual ability in chronic stroke patients: Rasch-based validation and relationship to upper limb impairment. Stroke, 32, 1627-34.
• Simone, A., Rota, V., Tesio, L., & Perucca, L. (2011). Generic ABILHAND questionnaire can measure manual ability across a variety of motor impairments. International Journal of Rehabilitation and Research, 34, 131-40.

See the measure

How to obtain the ABILHAND:

The ABILHAND is available in Penta, M., Tesio, L., Arnould, C., Zancan, A., & Thonnard, J-L. (2001). The ABILHAND questionnaire as a measure of manual ability in chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. patients: Rasch-based validation and relationship to upper limb impairment. StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., 32, 1627-34.

Purpose

The Action Research Arm Test (ARAT) is an evaluative measure to assess specific changes in limb function among individuals who sustained cortical damage resulting in hemiplegiaComplete paralysis of the arm, leg, and trunk on one side of the body that results from damage to the parts of the brain that control muscle movements. Hemiplegia is not a progressive condition, nor is it a disease. (Lyle, 1981). It assesses a client’s ability to handle objects differing in size, weight and shape and therefore can be considered to be an arm-specific measure of activity limitation (Platz, Pinkowski, Kim, di Bella, & Johnson, 2005).

In-Depth Review

Purpose of the measure

The Action Research Arm Test (ARAT) is an evaluative measure to assess specific changes in limb function among individuals who sustained cortical damage resulting in hemiplegiaComplete paralysis of the arm, leg, and trunk on one side of the body that results from damage to the parts of the brain that control muscle movements. Hemiplegia is not a progressive condition, nor is it a disease. (Lyle, 1981). It assesses a client’s ability to handle objects differing in size, weight and shape and therefore can be considered to be an arm-specific measure of activity limitation (Platz, Pinkowski, Kim, di Bella, & Johnson, 2005).

Available versions

The ARAT was developed by Ronald Lyle in 1981 by adapting the Upper Extremity Function Test (UEFT) (Carroll, 1965). The UEFT test administration and scoring was simplified, the time required to administer the test was shorted, and items were grouped based on the hierarchical scale (Guttman Scale) (Lang, Wagner, Dromerick, & Edwards, 2006). Due to the need for more specific and detailed instructions related to the client’s position, scoring and test administration, Yozbatiran, Der-Yeghiaian, and Cramer (2008) proposed a standardized approach to the ARAT.

Features of the measure

Items:

The ARAT consists of 19 items grouped into four subscales: grasp, grip, pinch, and gross movement. Each subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
constitutes a hierarchical Guttman scale, which means that all items are ordered according to ascending difficulty. In the ARAT, if the client succeeds in completing the most difficult item in a subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, this suggests he/she will succeed in the easier items for that same subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
. Similarly, failure on an item suggests the client will be unable to complete the remaining more challenging items in the subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
.

According to the rules defained by Lyle (1981), the client must first try to perform the most difficult task in a subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
. If the maximum score (score = 3) is obtained for this task then the maximum score for this entire subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
should be assigned, and the evaluator should move to the next subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
to be administered. When the client is unable to complete the most difficult item (scoring between 0-2), then the easiest item in this specific subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
should be performed. If the client fails completely (score = 0) when performing the easiest task, then the other intermediate items must not be tested, the entire subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
should be scored as zero, and the evaluator should then move to the next subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
. However, if the client succeeds at the easiest task either partially (score = 1 or 2) or completely (score = 3), then all the other tasks in that same subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
should be tested before moving to the next subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
. Following these rules, the items administered will range from a minimum of 4 to a maximum of 19 (van der Lee, Roorda, & Lankhorst, 2002).

The ARAT must be administered in a formal setting, since a specially designed table and chair are required (see equipment section for more information). For the starting position, the client should be seated in a chair, with a firm back and no armrests. The client’s trunk should be in contact with the back of the chair at all times during the test performance. Instructions about the required seating posture should be provided to the client prior to initiating the test. Additionally, reminders about the maintenance of this position should be given to the client when this condition is not respected. The client’s feet should be in contact with the floor throughout testing (van der Lee, DeGroot, Beckerman, Wagenaar, Lankhorst, & Bouter, 2001a; Yozbatiran et al., 2008). Both hands should be tested, beginning with the non- or less-affected hand, in order to practice and register baseline scores. Should the client be unable to understand the instructions for the required task, the evaluator should demonstrate the task and allow the client to try it as a trial (Yozbatiran et al., 2008). To facilitate recording the time for each task, the client’s hands should start and finish the task with palms down on the table. However, for the gross movement tasks, the client’s hands should be placed pronated on their lap. (Lyle, 1981; Yozbatiran et al., 2008).

In the grasp and pinch subscales, testing materials are lifted 37 cm from the surface of the table to the top of the shelf. In the grip subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, testing materials are moved from one side of the table to the other. Finally, in the gross movement subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, the client is requested to place the hand being tested either behind his/her head, on top of his/her head, or to his/her mouth (Lyle, 1981; Hsieh, Hsueh, Chiang, & Lin, 1998; Hsueh, Lee, & Hsieh, 2002a). The proper sequence for testing is 1) grasp subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, 2) grip subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, 3) pinch subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, 4) gross movement subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
(Lyle, 1981). The ARAT comes with simple instructions to guide the evaluator on scoring and administering the test (Lyle, 1981).

Scoring:

The ARAT is scored on a four-level ordinal scale (0-3) (Lyle, 1981).

• 0 = can not perform any part of the test,
• 1 = performs the test partially,
• 2 = completes the test, but takes abnormally long, time
• 3 = performs the test normally

In order to facilitate scoring, time limits have been suggested (Wagenaar, Meijer, van Wierinen, Kuik, Hazenberg, Lindeboom, Wichers, & Rijswijk, 1990; Yozbatiran et al., 2008). Incorporating the time limits to Lyle’s scoring definition, the new scoring system would be:

• 0 = cannot perform any part of the test;
• 1 = performs the test partially;
• 2 = completes the test, but takes an abnormally long time, varying from 5 to 60 seconds.

  If a client takes more than 60 seconds to perform an item, the evaluator should interrupt after 60 seconds and a score of 1 is given on that specific item.

• 3 = performs the test normally in less than 5 seconds.

The subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
scores range according to the number of items on each subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, as follows:

Subscales on the ARAT	Number of items per subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).	Score ranges per subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
Grasp subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).	6 items	Score 0-18
Grip subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).	4 items	Score 0-12
Pinch subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).	6 items	Score 0-18
Gross Movement subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).	3 items	Score 0-9

The total score on the ARAT ranges from 0 to 57, with the lowest score indicating that no movements can be performed, and the upper score indicating normal performance. Thus, higher scores will indicate better performance (Lang et al., 2006; van der Lee et al., 2002). The ARAT scores is a continuous measure, with no categorical cutoff scores. Therefore the score obtained at the ARAT does not allow classifying the clients into categories such as normal, mild limited, or severely limited.

Time:

The time required to complete the ARAT will depend on the number of items administered. Based on its hierarchical design, the ARAT was constructed to save testing time. Thus, no more than 7-10 minutes should be required to assess a client with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. (DeWeerdt, & Harrinson, 1985). However, if all 19 items are performed, the ARAT usually takes 20 minutes to administer (van der Lee et al., 2002). In one study by Hsieh and colleagues (1998), the ARAT took, on average, 8 minutes to administer to clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Subscales:

The ARAT is divided in four subscales: Grasp; Grip; Pinch and Gross movement.

The grasp and pinch subscales have 6 items each, the grip subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
has 4 items, and the gross movement has 3 items (Lyle, 1981).

Equipment:

Standardized equipment is required to administer the ARAT. It can be ordered only from Netherlands’ representatives. The average cost for this equipment is approximately 850 Euros ($1200 CAD) with an additional delivery fee of 179 Euros ($252 CAD).

The complete ARAT kit consists of:

• A specially designed table of 92cm x 45cm x 83cm high, with a shelf of 93cm x 10cm, positioned 37cm above the main surface of the table (Lyle, 1981; Hsueh et al., 2002a).
• A chair with back rest and no arm rests, that should be placed 44cm above floor level (Lyle, 1981; Hsueh et al., 2002a).
• Woodblocks of 2.5, 5, 7.5 and 10cm³ (Lyle, 1981; Hsueh et al., 2002a).
• A cricket ball 7.5cm in diameter (Lyle, 1981; Hsueh et al., 2002a).
• Two alloy tubes: one 2.25cm in diameter x 11.5 cm long, the second one 1.0cm in diameter x 16cm long (Lyle, 1981; Hsueh et al., 2002a).
• A washer and bolt; which is a type of screw with its anchor (Lyle, 1981; Hsueh et al., 2002a).
• Two glasses (Lyle, 1981; Hsueh et al., 2002a).
• A marble 1.5cm in diameter (Lyle, 1981; Hsueh et al., 2002a).
• A ball bearing 6mm in diameter (Lyle, 1981; Hsueh et al., 2002a).
• A stopwatch (Wagenaar et al., 1990; Yozbatiran et al., 2008)
• Paper and pencil for the evaluator.

Training:

None typically reported.

Alternative forms of the Action Research Arm Test

None.

Client suitability

Can be used with:

• The ARAT was constructed for assessing recovery of upper limb function following cortical damage (Lyle, 1981).
• Clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Should not be used in:

• When administering the ARAT for clients with finger amputation, pinch subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
  should be scored as 0 as well all other tasks that require movement of an amputated body part (Yozbatiran et al., 2008).

In what languages is the measure available?

There are no official translations of the ARAT.

Nevertheless, some peer-reviewed publications from the Netherlands and Taiwan have used the ARAT as an outcome measure, which may indicate that instructions have been informally translated to other languages (Hsieh et al., 1998; Hsueh et al., 2002a; van der Lee et al., 2002).

Summary

What does the tool measure?	The ARAT measures specific changes in limb function among individuals who sustained cortical damage resulting in hemiplegiaComplete paralysis of the arm, leg, and trunk on one side of the body that results from damage to the parts of the brain that control muscle movements. Hemiplegia is not a progressive condition, nor is it a disease..
What types of clients can the tool be used for?	The ARAT can be used with, but is not limited to clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
Time to administer	An average of 7 to 10 minutes.
Versions	There are no alternative versions.
Other Languages	There are no official translations.
Measurement Properties
ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: One study examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the ARAT and reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. using Cronbach’s alpha. Test-retest: Three studies have examined the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of the ARAT. All reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). using ICCs. Intra-rater: Four studies have examined the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the ARAT and reported excellent intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. using Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. , intraclass correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. coefficients (ICC) and weighted kappa. Inter-rater: Seven studies examined the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the ARAT and reported excellent inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. using Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. , Intra ICC and weighted kappa.
ValidityThe degree to which an assessment measures what it is supposed to measure.	Criterion: Concurrent: One study has examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.” of the ARAT and reported adequate to excellent correlations with the Box and Block Test (BBT) and the Nine-Hole Peg Test (NHPT) at pre and post-treatment. Predictive: No studies have examined the predictive validityA form of criterion validity that examines a measure’s ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. of the ARAT. Construct: Convergent: Seven studies examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other. of the ARAT and reported excellent correlations between the ARAT and the Brunnstrom-Fugl-Meyer test; the upper extremity subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society). of the Motor Assessment scale; the Motricity Index; the upper extremity movement of Modified Motor Assessment Chart; the BTT; the motor function subscore of the Fugl-Meyer test; the Hemispheric StrokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. Scale; upper extremity strength and grasp speed. Adequate correlations were reported between the ARAT and the passive joint motion/joint pain of the Fugl-Meyer test, the Functional Independence Measure and spasticityInvoluntary muscle tightness and stiffness that can occur after a stroke. It is characterized by exaggerated deep tendon reflexes that interfere with muscular activity, gait, movement, or speech. Spasticity can increase initially but wane down later on, after stroke. . Poor correlations were reported between the ARAT and the sensation score of the Fugl-Meyer test, the Ashworth scale, the Modified Barthel Index, the National Institutes of Health StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale, the light touch sensation and pain.
Floor/Ceiling Effects	– One study examined the floor/ceiling effects of the ARAT in clients with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and reported that at earlier phases of the strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., floor effects were poor. At discharge from the acute rehabilitation ward, ceiling effects on the ARAT were adequate. – One study examined the floor/ceiling effects of the ARAT in strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. clients with mild to moderate hemiparesis and reported adequate floor and ceiling effects.
SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” / SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).	No studies have examined the specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative). of the ARAT.
Does the tool detect change in patients?	Six studies have examined the responsivenessThe ability of an instrument to detect clinically important change over time. of the ARAT and reported that the ARAT has a moderate to large Standardized Response MeanThe standardized response mean (SRM) is calculated by dividing the mean change by the standard deviation of the change scores. , moderate to large effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”. and large responsivenessThe ability of an instrument to detect clinically important change over time. ratio, therefore, is able to detect change in clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Acceptability	When administering the ARAT to clients with upper extremity amputations attention is required when scoring (i.e. – a score of 0 is given).
Feasibility	The administration of the ARAT is quick and simple, but requires standardized equipment.
How to obtain the tool?	Information on the ARAT can be obtained in the study by Lyle (1981), Hsieh et al. (1998), van der Lee et al. (2002), Rabadi & Rabadi (2006), and Yozbatiran et al. (2008) and at the website: http://www.aratest.eu/Index_english.htm Standardized equipment can be purchased from the following website: http://www.aratest.eu/ or from http://www.saliarehab.com/

Psychometric Properties

Overview

We conducted a literature search to identify all relevant publications on the psychometric properties of the Action Research Arm Test (ARAT) in individuals with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. We identified twelve studies. The ARAT appears to be floor effects.

Floor/Ceiling Effects

Hsueh and Hsieh (2002b) examined floor and ceilings effects for the ARAT and the Upper Extremity Motor Assessment Scale (Carr, Shepherd, Nordholm, & Lynne, 1985) in 48 clients with acute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were assessed at admission and discharge from an acute rehabilitation ward. At admission, the ARAT total score demonstrated a poor floor effectThe floor effect is when data cannot take on a value lower than some particular number. Thus, it represents a subsample for whom clinical decline may not register as a change in score, even if there is worsening of function/behavior etc. because there are no items or scaling within the test that measure decline from the lowest possible score. See also "ceiling effect."
, with 52.1% of participants scoring 0. Although all subscales were classified as having a poor floor effectThe floor effect is when data cannot take on a value lower than some particular number. Thus, it represents a subsample for whom clinical decline may not register as a change in score, even if there is worsening of function/behavior etc. because there are no items or scaling within the test that measure decline from the lowest possible score. See also "ceiling effect."
, when comparing ARAT’s subscales among themselves, 72.9% of participants were unable to perform the pinch subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
, 70.8% were unable to perform both grasp and grip subscales and 52.1 % were unable to complete the gross movement subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
. At discharge, the ARAT total score demonstrated an adequate ceiling effectA ceiling effect occurs when test items aren't challenging enough for a group of individuals. Thus, the test score will not increase for a subsample of people who may have clinically improved because they have already reached the highest score that can be achieved on that test. In other words, because the test has a limited number of difficult items, the most highly functioning individuals will score at the highest possible score. This becomes a measurement problem when you are trying to identify changes - the person may continue to improve but the test does not capture that improvement. Example: A memory test that assesses how many words a participant can recall has a total of five words that each participant is asked to remember. Because most individuals can remember all five words, this measure has a ceiling effect. See also "floor effect.", with only 7% of participants scoring the maximal value. When analyzing ARAT’s subscales individually the gross movement subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
presented the poorest ceiling effectA ceiling effect occurs when test items aren't challenging enough for a group of individuals. Thus, the test score will not increase for a subsample of people who may have clinically improved because they have already reached the highest score that can be achieved on that test. In other words, because the test has a limited number of difficult items, the most highly functioning individuals will score at the highest possible score. This becomes a measurement problem when you are trying to identify changes - the person may continue to improve but the test does not capture that improvement. Example: A memory test that assesses how many words a participant can recall has a total of five words that each participant is asked to remember. Because most individuals can remember all five words, this measure has a ceiling effect. See also "floor effect.", with 29.2% of participants scoring the maximum score, followed by 27% of participants on the grasp subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
. The grip and pinch subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
had the best classification, with an adequate ceiling effectA ceiling effect occurs when test items aren't challenging enough for a group of individuals. Thus, the test score will not increase for a subsample of people who may have clinically improved because they have already reached the highest score that can be achieved on that test. In other words, because the test has a limited number of difficult items, the most highly functioning individuals will score at the highest possible score. This becomes a measurement problem when you are trying to identify changes - the person may continue to improve but the test does not capture that improvement. Example: A memory test that assesses how many words a participant can recall has a total of five words that each participant is asked to remember. Because most individuals can remember all five words, this measure has a ceiling effect. See also "floor effect." of 18.8% and 16.7%, respectively.

Compared to the ARAT, at admission the Upper Extremity Motor Assessment Scale had 58% of participants scoring the minimal value, indicating a poor floor effect. However, at discharge the Upper Extremity Motor Assessment Scale demonstrated a more adequate ceiling effectA ceiling effect occurs when test items aren't challenging enough for a group of individuals. Thus, the test score will not increase for a subsample of people who may have clinically improved because they have already reached the highest score that can be achieved on that test. In other words, because the test has a limited number of difficult items, the most highly functioning individuals will score at the highest possible score. This becomes a measurement problem when you are trying to identify changes - the person may continue to improve but the test does not capture that improvement. Example: A memory test that assesses how many words a participant can recall has a total of five words that each participant is asked to remember. Because most individuals can remember all five words, this measure has a ceiling effect. See also "floor effect." than the ARAT, with only 4.3 % of participants obtaining the maximum score.

Nijland et al. (2010) investigated the psychometric properties of the ARAT and Wolf Motor Function Test in 40 patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. with mild to moderate hemiparesis. The ARAT showed adequate floor and ceiling effects with only 12.5 to 17% of patients scoring the lowest or highest scores.

Reliability

Internal ConsistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.:
Nijland et al. (2010) investigated the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the ARAT in 40 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. with mild to moderate hemiparesis. Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the ARAT, as calculated using Cronbach’s Coefficient Alpha was excellent (α = 0.98).

Test-retest:
Note: From the descriptions provided of the following studies it appears that some authors called the testing test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
while others called the same analysis intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
.

Lyle (1981) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
in 20 individuals who sustained cortical damage, either from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. or traumatic brain lesion. The mean age was 53 years, ranging from 26 to 72 years. Participants were re-assessed with a 1-week interval by the same rater and under the same conditions. The test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
, as calculated using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (r = 0.98).

Hsueh, Lee, and Hsieh (2002a) evaluated test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
performed using a regular table instead of the specially designed table for this test in 61 individuals with sub-acute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. and a mean age of 63 years old. Participants were re-assessed after a two-day interval by the same rater. The test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
, as calculated using the Intraclass Correlation Coefficient (ICC)Intraclass correlation (ICC) is used to measure inter-rater reliability for two or more raters. It may also be used to assess test-retest reliability. ICC may be conceptualized as the ratio of between-groups variance to total variance., was excellent for the total score (ICC = 0.99) as well as for the grasp, grip, pinch and gross movement subscales (ICC = 0.99, 0.98, 0.96 and 0.95, respectively).

Platz, Pinkowski, van Wijck, Kim, di Bella, and Johnson (2005) estimated test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
for the ARAT, the Box and Block Test (Cromwell, 1965; Mathiowetz, Volland, Kashman, & Weber, 1985a), and the Fugl-Meyer Test upper extremity items (including items from the Motor function, Sensation and Passive Joint Motion/Joint pain subscores) (Fugl-Meyer, Jääskö, Leyman, Olsson, & Steglind, 1975) in 23 participants with upper extremity paresis either from strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, or traumatic brain injury. The participant’s most affected arm was re-assessed 1 week later by the same rater. The test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the ARAT total score, as calculated using ICC’s and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.96 and rho = 0.96). Furthermore, test-retest reliabilities for each subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
were all excellent: grasp (ICC = 0.94 and rho = 0.96), grip (ICC = 0.94 and rho = 0.95), pinch (ICC = 0.89 and rho = 0.89) and gross movement (ICC = 0.97 and rho = 0.97).
Note: These results applies only to the most affected upper limb.

Intra-rater:
Wagenaar, Meijer, van Wierinen, Kuik, Hazenberg, Lindeboom, Wichers and Rijswijk (1990) evaluated intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
in seven patients with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. The timeframe for assessments were not provided by the author. Intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
as calculated using Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (rho = 0.99).

Van der Lee, DeGroot, Beckerman, Wagenaar, Lankhorst, and Bouter (2001a) estimated intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
in 20 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and a median age of 62 years. Participants were evaluated by the same rater at three points in time. At the baseline assessment participants were videotaped. The second assessment was 4-27 months following the first assessment, and the final assessment was 4-6 weeks after. Scoring the last two assessments was based on the videotaped recorded at baseline. Intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
results were analyzed between the two first assessments, where scoring sources were different (live vs. videotape) and between the two last assessments, were scoring sources were the same (videotape only). Intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
, as calculated using ICC and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.99 and rho = 0.99), independent of scoring sources. Intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
, as calculated using weighted kappa was also excellent: scoring with the same information source resulted in a kappa = 1.00 versus only a slightly lower kappa when scoring from two different information sources (kappa = 0.94). The gross movement subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
showed the lowest weighted kappa value (kappa = 0.83), suggesting that this subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
had the lowest agreement level.

Yozbatiran, Der-Yeghiaian, and Cramer (2008) examined intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
in 8 clients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were re-assessed by the same rater and under the same conditions with a 1-week interval. Intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
for the total score, as calculated using ICC and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.99 and rho = 0.99). Additionally, the same excellent level of intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
was found for the grasp, grip, pinch, and gross motor movement subscales (ICC = 0.98 and rho = 0.93; ICC = 0.97 and rho = 0.93; ICC = 0.99 and rho = 0.98; ICC = 0.93 and rho = 0.91, respectively).

Nijland et al. (2010) investigated the psychometric properties of the ARAT and Wolf Motor Function Test in 40 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. with mild to moderate hemiparesis. 18 patients participated in the reproducibility testing of the ARAT and were assessed twice by the same observer approximately 10 days apart. Intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
, as analyzed using the ICC was found to be excellent (ICC = 0.97).

Inter-rater:
Lyle (1981) examined inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
in 20 individuals who had sustained cortical damage, either from strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. or traumatic brain injury. The mean age was 53 years, ranging from 26 to 72 years. Participants were assessed independently by two different raters. Agreement between raters as calculated using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (r = 0.99).

Hsieh, Hsueh, Chiang, and Lin (1998) assessed inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
in 50 clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Their mean age was 65 years old. Participants were evaluated independently, on three different days, by three raters. ICC for the total score showed excellent agreement (ICC = 0.98). Agreement between raters was also excellent for grasp, grip, pinch and gross movement subscales (ICC = 0.98; ICC = 0.96; ICC = 0.96; ICC = 0.95, respectively).

Van der Lee et al. (2001a) estimated inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
in 20 patients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. and a median age of 62 years old. Participants were videotaped and scored independently by two raters. Inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
, as calculated using ICC, weighted kappa, and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.98; kappa = 0.93; rho = 0.99). With respect to the individual subscales, the gross movement scale had the lowest weighted kappa value (kappa = 0.87), suggesting this subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
has the lowest agreement between raters.

Hsueh, Lee, and Hsieh (2002a) evaluated inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the ARAT performed with a regular table instead of the specially designed table for this test in 61 individuals with sub-acute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. and a mean age of 63 years old. Participants were re-assessed with a two-day interval by three different raters. ICC for the total score showed excellent agreement (ICC = 0.99) as well as for grasp, grip, pinch and gross movement subscales (ICC = 0.99; ICC = 0.98; ICC = 0.96; ICC = 0.94, respectively).

Platz et al. (2005) analyzed inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the ARAT, the Box and Block Test and the Fugl-Meyer Test upper extremity items (including items from the Motor function, Sensation and Passive Joint Motion/Joint pain subscores) in 44 individuals with upper limb paresis either from strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, or traumatic brain injury. Participants had the most affected arm videotaped and scored independently by two raters. Inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
for the ARAT total score, as calculated using the ICC and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.99 and rho = 0.99). Additionally, the scores for each subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
were provided and inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
for grasp (ICC = 0.99 and rho = 0.99), grip (ICC = 0.96 and rho = 0.95), pinch (ICC = 0.99 and rho = 0.99) and gross movement (ICC = 0.98 and rho = 0.98) subscales were all excellent.
Note: These results applies only to the most affected upper limb.

Yozbatiran et al. (2008) evaluated inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
in 9 clients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were scored simultaneously and independently by two raters. Inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
for the total score, as calculated using the ICC and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.99 and rho = 0.96). The same excellent level of inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
was found for the grasp, grip, pinch and gross motor movement subscales (ICC = 0.99 and rho = 1; ICC = 0.99 and rho = 0.99; ICC = 0.99 and rho = 0.98; ICC = 0.97 and rho = 0.93, respectively).

Nijland et al. (2010) investigated the psychometric properties of the ARAT and Wolf Motor Function Test in 40 patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. with mild to moderate hemiparesis. 18 patients participated in the reproducibility testing of the ARAT and were assessed in random order by two observers, within one week. Inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
, as analyzed using the ICC was found to be excellent (ICC = 0.92).

Validity

Content:

Lyle, 1981 generated the 19 ARAT items from the 33 items of the Upper Extremity Function Test (UEFT – Caroll, 1965). Item reduction was based on a low inter-item correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, on item redundancy, confirmed through a very high inter-item correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
(above r = 0.9) and on items that were extremely difficult to perform. Nevertheless, ARAT items were not based on a theoretical model (Finch, Brooks, Stratford, & Mayo, 2002).

Criterion:

Concurrent:
No gold standardA measurement that is widely accepted as being the best available to measure a construct.
exists against which to compare the ARAT.

Lin, Chuang, Wu, Hsieh and Chang (2010) compared the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the ARAT, Box and Block Test (BBT) and Nine-Hole Peg Test (NHPT) for evaluating hand dexterity in 59 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. The Fugl-Meyer Assessment of Sensorimotor Recovery After StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (FMA), Motor Activity Log (MAL) and StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale (SIS) were also administered to assess the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the ARAT, BBT and NHPT. Using Spearman rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient, the ARAT, BBT and NHPT were found to have adequate to excellent correlations at pre-treatment (ranging from rho=-0.55 to -0.80) and post-treatment (ranging from rho=-0.57 to -0.71). In addition, the ARAT and BBT were found to have adequate correlations with the FMA, MAL and SIS (ranging from rho=0.31-59); however, the NHPT had only poor to adequate correlations with the FMA and MAL (ranging from rho=-0.16 to -0.33); and adequate to excellent correlations with the SIS (ranging from rho=-0.58 to -0.66). When considering both the results of responsivenessThe ability of an instrument to detect clinically important change over time.
and validation components of the study, the ARAT and BBT are believed to be more appropriate than the NHPT for evaluating dexterity.

Predictive:
No studies have examined the predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the ARAT.

Construct:

Convergent/Discriminant:
DeWeerdt and Harrison (1985) evaluated the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT by comparing it to the Fugl-Meyer test (Fugl-Meyer et al., 1975) in 53 clients with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Their mean age was 68 years. Correlations were calculated at two points in time after strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. onset using Spearman correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Excellent correlations were found between the ARAT and Fugl-Meyer test at 2 months (rho = 0.91) and at 8 months (rho = 0.94) post-stroke.

Wagenaar, Meijer, van Wierinen, Kuik, Hazenberg, Lindeboom, Wichers and Rijswijk (1990) evaluated the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT by comparing it to the Sollerman test (Jacobson-Sollerman & Sperling, 1977) in seven patients with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. An excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, as calculated using Spearman rho, was found (rho = 0.94).
Note: The Sollerman test measures hand grip function using 20 different daily life activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
requiring hand movements.

Hsieh et al. (1998) assessed convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT by comparing it to the Upper Extremity portion of the Motor Assessment Scale (Carr et al., 1985), the arm subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
of the Motricity Index (Demeurisse, Demol, & obaye, 1980), and the upper extremity movements of the Modified Motor Assessment Chart (Lindmark & Hamrin, 1988) in 50 clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. The mean age of clients was 65 years old. Correlations were calculated using Pearson CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
Coefficients. Excellent correlations were found between the ARAT and the Upper Extremity part of the Motor Assessment Scale ((r = 0.96), Motricity Index (r = 0.87) and the upper extremity movements of the Modified Motor Assessment Chart (r = 0.94).

Platz et al. (2005) tested convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT by comparing it to the Box and Block Test (Cromwell, 1965; Mathiowetz et al., 1985a), the Fugl-Meyer Test upper extremity items (including items from the Motor Function, Sensation and Passive Joint Motion/Joint Pain subscores) (Fugl-Meyer et al., 1975), the Motricity Index (Demeurisse et al., 1980), the Ashworth Scale (Ashworth, 1964), the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale (Adams, Meador, Sethi, Grotta, & Thomson, 1986) and the Modified Barthel Index (Collin, Wade, Davies, & Horne, 1988) in 56 participants with upper extremity paresis either from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=37), multiple sclerosis (n=14), or traumatic brain injury (n=5). Correlations were calculated using the Spearman CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
Coefficient. Excellent correlations were found between the ARAT and the Box and Block Test (rho = 0.95), the Motor Function subscore of the Fugl-Meyer Test (rho = 0.92), the Motricity Index (rho = 0.81), and the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale (rho = -0.66). Adequate correlations were found between the ARAT and the Passive Joint Motion/Joint Pain subscore of Fugl Meyer Test (rho = 0.42). Poor correlations were found between the ARAT and the Sensation Subscore of the Fugl-Meyer Test (rho = 0.29), the Ashworth Scale (rho = -0.29) and the Modified Barthel Index (rho = 0.04).
Note: Negative correlations are observed because a high score on the ARAT indicates normal performance, whereas a low score on the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale and the Ashworth Scale indicates normal performance.

Lang, Wagner, Dromerick, and Edwards (2006) evaluated the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT in 50 individuals with acute to sub acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., mean age of 63 years old, attending an acute neurology strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. service at three points in time: admission (day 0); post intervention (day 14); and 90 days poststroke (day 90). The ARAT was compared to measures of sensorimotor impairment (e.g. light touch sensation, pain, elbow joint spasticityInvoluntary muscle tightness and stiffness that can occur after a stroke. It is characterized by exaggerated deep tendon reflexes that interfere with muscular activity, gait, movement, or speech. Spasticity can increase initially but wane down later on, after stroke.
, upper extremity strength), to kinematic measures (e.g. reach and grasp), to the Functional Independence Measure (FIM) (Keith, Granger, Hamilton, & Sherwin, 1987), and to the National Institutes of Health StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale (NIHSS) (Brott, Adams, Olinger, Marler, Barsan, Biller, et al., 1989). At day 0, excellent correlations were found between the ARAT and upper extremity strength (r = 0.60) and grasp speed (r = 0.60). Adequate correlations were found between the ARAT and grasp efficiency (r = 0.42), reach efficiency (r = -0.38) and reach speed (r = 0.40), and the FIM upper extremity score (r = 0.38). Poor correlations were found between the ARAT and NIHSS (r = -0.15); light touch sensation (r = 0.15), pain (r = 0.10), elbow joint spasticityInvoluntary muscle tightness and stiffness that can occur after a stroke. It is characterized by exaggerated deep tendon reflexes that interfere with muscular activity, gait, movement, or speech. Spasticity can increase initially but wane down later on, after stroke.
(r = -0.28) and the FIM total score (r = 0.20). At day 14, excellent correlations were found between the ARAT and grasp efficiency (r = 0.60) and the FIM upper extremity scores (r = 0.62). Adequate correlations were found between the ARAT and elbow spasticityInvoluntary muscle tightness and stiffness that can occur after a stroke. It is characterized by exaggerated deep tendon reflexes that interfere with muscular activity, gait, movement, or speech. Spasticity can increase initially but wane down later on, after stroke.
(r = 0.49), upper extremity strength (r = 0.42), reach efficiency (r = -0.58), grasp speed (r = 0.36) and the FIM total score (r = 0.52). Poor correlations were found between the ARAT and NIHSS (r = -0.24), light touch sensation (r = -0.20), and pain (r = -0.12). At day 90, excellent correlations were found between the ARAT and upper extremity strength (r = 0.60). Adequate correlations were found between the ARAT and elbow spasticityInvoluntary muscle tightness and stiffness that can occur after a stroke. It is characterized by exaggerated deep tendon reflexes that interfere with muscular activity, gait, movement, or speech. Spasticity can increase initially but wane down later on, after stroke.
(r = -0.42), reach efficiency (r = -0.42), reach speed (r = 0.50), grasp efficiency (r = -0.48), grasp speed (r = 0.38) and the FIM upper extremity (r = 0.42) and total scores (r = 0.40). Poor correlations were found between the ARAT and the NIHSS (r = -0.29), light touch sensation (r = 0.00), and pain (r = 0.22). In summary, from this study’s findings it appears that the NIHSS, light touch sensation, and pain do not appear to relate to the ARAT. The relationship between the ARAT and FIM scores is stronger early on post-stroke and stabilizes by the ninetieth day.

Rabadi and Rabadi (2006) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT by comparing it to the Fugl-Meyer Assessment (Fugl-Meyer et al., 1975) at admission and discharge from an acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. rehabilitation unit in 104 inpatients with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. with a mean age of 72 years. The correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between ARAT and the Fugl-Meyer Assessment was excellent both at admission (rho = 0.77) and discharge (rho = 0.87).

Yozbatiran et al. (2008) estimated the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the ARAT by comparing it to the arm motor Fugl-Meyer Assessment (Fugl-Meyer et al., 1975) score in 12 clients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. at a mean age of 61 years. Excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
(r = 0.94) was found between the ARAT and arm motor Fugl-Meyer score.

Known groups:
No studies have examined known groups validityKnown groups validity is a form of construct validation in which the validity is determined by the degree to which an instrument can demonstate different scores for groups know to vary on the variables being measured.
of the ARAT.

Responsiveness

Van der Lee, Beckerman, Lankhorst, and Bouter (2001b) evaluated the responsivenessThe ability of an instrument to detect clinically important change over time.
on the ARAT and Fugl-Meyer Assessment (Fugl-Meyer et al., 1975) in 22 clients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., mean age of 58 years old, receiving intensive forced use treatment. Participants were assessed two weeks pre- and two weeks post- treatment. A responsivenessThe ability of an instrument to detect clinically important change over time.
ratio was calculated. Compared to the Fugl-Meyer Assessment, the ARAT had a greater responsivenessThe ability of an instrument to detect clinically important change over time.
ratio (2.03 for ARAT vs. 0.41 for Fugl-Meyer) suggesting that the ARAT is more sensitive to detecting change.
Note: The responsivenessThe ability of an instrument to detect clinically important change over time.
ratio is a variant of effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
and higher values indicate better responsivenessThe ability of an instrument to detect clinically important change over time.
.

Van der Lee, Roorda, Beckerman, and Lankhorst (2002) estimated the responsivenessThe ability of an instrument to detect clinically important change over time.
of a modified version of the ARAT in 63 participants with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. In this study, researchers did not follow Lyle’s standardized instructions. Instead, they administered all 19 ARAT items to verify any possible effect of this format on its psychometric properties. A responsivenessThe ability of an instrument to detect clinically important change over time.
ratio was calculated. Compared to the hierarchical version proposed by Lyle, performing all 19 items was found to improve the measure’s responsivenessThe ability of an instrument to detect clinically important change over time.
, with a responsivenessThe ability of an instrument to detect clinically important change over time.
ratio of 1.7 compared to 1.2 with Lyle’s version.
Note: The responsivenessThe ability of an instrument to detect clinically important change over time.
ratio can be considered an estimate of effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
normalized to the variability in a stable population and higher values indicate better responsivenessThe ability of an instrument to detect clinically important change over time.
.

Hsueh et al. (2002b) analyzed the responsivenessThe ability of an instrument to detect clinically important change over time.
of the ARAT and the upper extremity section of the Motor Assessment Scale (Carr et al., 1985) in 48 participants having acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and a mean age of 62 years. Participants were assessed at two points in time: admission and discharge from the acute rehabilitation centre. The ARAT total score demonstrated a moderate effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
of 0.52, while the Motor Assessment Scale total score demonstrated a small effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
of 0.45.

Lang et al. (2006) examined the responsivenessThe ability of an instrument to detect clinically important change over time.
of the ARAT in 50 participants with acute to subacute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., with a mean age of 63 years old, receiving constraint-induced movement therapy (CIMT). Assessments were performed at three points in time: baseline, immediately post-treatment, and 2.5 months post-treatment. Effects sizes and responsivenessThe ability of an instrument to detect clinically important change over time.
ratios were calculated. ARAT total and subscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
scores at the first follow-up evaluation were similar, with moderate to large effect sizes (ARAT total score = 1.01; grasp subscore = 1.04; pinch subscore = 0.85; grip subscore = 1.01; and gross movement subscore = 0.72). The second follow-up evaluation demonstrated large effect sizes, with individual higher values when compared to the first evaluation (ARAT total score = 1.39; grasp subscore = 1.22; pinch subscore = 1.49; grip subscore = 1.32 and gross movement subscore = 0.98). The responsivenessThe ability of an instrument to detect clinically important change over time.
ratio for the ARAT total score at the first follow-up evaluation was 5.2 and at the second was 7.0. These two responsivenessThe ability of an instrument to detect clinically important change over time.
estimations suggest that the ARAT is a sensitive tool for detecting change even months after strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. onset.
Note: ResponsivenessThe ability of an instrument to detect clinically important change over time.
ratio is a variant of effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
and higher values indicate better responsivenessThe ability of an instrument to detect clinically important change over time.
.

Rabadi and Rabadi (2008) assessed the responsivenessThe ability of an instrument to detect clinically important change over time.
of the ARAT and the Fugl-Meyer Assessment (Fugl-Meyer et al., 1975) in 104 participants with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., with a mean age of 72 years, undergoing inpatient rehabilitation. Participants were evaluated at admission and discharge from acute care. The Standardized Response MeanThe standardized response mean (SRM) is calculated by dividing the mean change by the standard deviation of the change scores.
(SRM) was used to calculate responsivenessThe ability of an instrument to detect clinically important change over time.
. Amongst these upper extremity tests, the ARAT was less sensitive than the Fugl-Meyer Assessment (SRM = 0.68 and 0.74, respectively). However, since the difference between the SRMs for these two measures was minimal, these tests can be considered equally sensitive to change during inpatient acute rehabilitation. This result is contrary to the one presented by Van der Lee at al. (2002). The reason for this difference may be due to the difference in these studies population age and strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. severity.
Note: SRM is a variant of effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
and higher values indicate better responsivenessThe ability of an instrument to detect clinically important change over time.
.

Lin, Chuang, Wu, Hsieh and Chang (2010) evaluated the responsivenessThe ability of an instrument to detect clinically important change over time.
of the ARAT, Box and Block Test (BBT), the Nine-Hole Peg Test (NHPT) for evaluating hand dexterity in 59 patients with subacute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (< 6-months) and Brunnstrom stage IV to VI for proximal and distal upper extremity function. Patients were randomly assigned to receive constraint-induced therapyA form of intervention that involves restraining the unaffected upper or lower extremity in order to encourage movement of the affected limbs. For persons with USN, constraint-induced therapy involves restraining the unaffected arm or hand using a sling or padded mitt, in order to promote visual scanning and movement in the neglected hemispace.
, bilateral arm training or control treatment and received 2 hours of therapy, 5 days per week for 3 weeks. Assessments were performed at baseline and 3 weeks. Using Standardized Response MeanThe standardized response mean (SRM) is calculated by dividing the mean change by the standard deviation of the change scores.
(SRM) to calculate responsivenessThe ability of an instrument to detect clinically important change over time.
, the ARAT, BBT and NHPT were all found to have moderate SRM (0.79, 0.74, 0.64 respectively), indicating sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
for detecting change in hand dexterity. When considering both the results of responsivenessThe ability of an instrument to detect clinically important change over time.
and validation components of the study, the ARAT and BBT are believed to be more appropriate than the NHPT for evaluating dexterity.

References

• Adams, R.J., Meador, K.J., Sethi, K.D., Grotta, J.C., & Thomson, D.S. (1986). Graded neurologic scale for the use in acute hemispheric stroke treatment protocols. Stroke, 18, 665-669.
• Ashworth, B. (1964). Preliminary trial of carisoprodol in multiple sclerosis. Practitioner, 192, 540-542.
• Brott, T. G., Adams, H. P., Olinger, C. P., Marler, J. R., Barsan, W. G., Biller, J., Spilker, J., Holleran, R., Eberle, R., Hertzberg, V., Rorick, M., Moomaw, C. J., & Walker, M. (1989). Measurements of acute cerebral infarction: a clinical examination scale. Stroke, 20, 864 -70.
• Carroll, D. (1965). A quantitative test of upper extremity function. Journal of Chronic Disability, 18, 479-91.
• Carr, J.H., Shepherd, R.B., Nordholm, L., & Lynne, D. (1985). Investigation of a new motor assessment scale for stroke patients. Physical Therapy, 65, 175- 180.
• Collin, C., Wade, D.T., Davies, S., & Horne, V. (1988). The Barthel ADL Index: a reliability study. International Disability Study, 10, 61-63.
• Cromwell, F.S (1965). Occupational therapists manual for basic skills assessment: primary prevocational evaluation. Pasadena, (CA): Fair Oaks Printing; 29-31.
• Demeurisse, G., Demol, O., & Robaye, E. (1980). Motor evaluation in vascular hemiplegia. European Neurology, 19(6), 382-389.
• De Weerdt, W.J.G., & Harrison, M.A. (1985). Measuring recovery of arm hand function in stroke patients: a comparison of the Brunnstrom-Fugl-Meyer test and the Action Research Arm Test. Physiotherapy Canada, 37, 65-70.
• Finch, E., Brooks, D., Stratford,P.W, & Mayo, N.E. (2002). Physical Outcome Measures: A guide to enhance physical outcome measures. Ontario, Canada: Lippincott, Williams, & Wilkins.
• Fugl-Meyer, A.R., Jääskö, L., Leyman, I., Olsson, S., & Steglind, S. (1975). The post-stroke hemiplegic patient 1. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine, 7, 13-31.
• Gowland, C., Van-Hullenaar, S., Torresin, W., et al., (1995). Chedoke-McMaster Stroke Assessment: development, validation, and administration manual. Hamilton, (ON), Canada: School of Rehabilitation Science, McMaster University
• Heller, A., Wade, D.T., Wood, V.A., Sunderland, A., Hewer, R., & Ward, E. (1987). Arm function after stroke: measurement and recovery over the first three months. Journal of Neurology, Neurosurgery & Psychiatry, 50(6), 714-719.
• Hsieh, C.L., Hsueh, I.P, Chiang, F., & Lin, P. (1998). Inter-rater reliability and validity of the action research arm test in stroke patients. Age and Ageing, 27, 107-113.
• Hsueh, I.P, Lee, M.M., & Hsieh, C.L. (2002a). The action research arm test: Is it necessary for patients being tested to sit at a standardized table? Clinical Rehabilitation, 16, 382-388.
• Hsueh, I.P. & Hsieh, C.L. (2002b). Responsiveness of two upper extremity function instruments for stroke inpatients receiving rehabilitation. Clinical Rehabilitation, 16, 617-624.
• Jacobson-Sollerman, X & Sperling, Y. (1977). Grip function of the healthy hand in a standardized hand function test. A study of the Rancho Los Amigos test. Scandinavian Journal of Rehabilitation Medicine, 9(3), 123-129.
• Keith, R.A, Granger, C.V., Hamilton, B.B., & Sherwin, F.S. (1987). The Functional Independence Measure: a new tool for rehabilitation. In: Eisenberg, M.G. & Grzesiak, R.C. (Ed.), Advances in clinical rehabilitation (pp. 6-18). New York: Springer Publishing Company.
• Kellor, M., Frost, J., Silberberg, N., Iversen, I., & Cummings R. (1971). Hand strength and dexterity. American Journal of Occupational Therapy, 25, 77-83.
• Lang, C.E., Wagner, J.M, Dromerick, A.W., & Edwards, D.F. (2006). Measurement of upper extremity function early after stroke: properties of the action research arm test. Archives Physical Medicine and Rehabilitation, 87, 1605-1610.
• Lin, K-C., Chuang, L-L., Wu, C-Y., Hseih, Y-W. & Chang, W-Y. (2010). Responsiveness and validity of three dexterous function measures in stroke rehabilitation. Journal of Rehabilitation Research and Development, 47(6), 563-572.
• Lindmark, B. & Hamrin, E. (1988). Evaluation of function capacity after stroke as a basis for active intervention: Presentation of a modified chart for motor capacity assessment and its reliability. Scandinavian Journal of Rehabilitation Medicine, 20, 103-109.
• Lyle, R.C. (1981). A performance test for assessment of upper limb function in physical rehabilitation treatment and research. International Journal of Rehabilitation and Research, 4, 483-492.
• Mathiowetz, V., Volland, G., Kashman, N., & Weber, K. (1985a). Adult norms for the box and block test of manual dexterity. American Journal of Occupational Therapy, 39, 386-391.
• Mathiowetz, V., Weber, K., Kashman, N., & Volland, G. (1985b). Adult norms for the nine hole peg test of finger dexterity. Occupational Therapy Journal of Research, 5, 24 -33.
• Nijland, R., van Wegen, E., Verbunt, J, van Wijk, R., van Kordelaar, J. & Kwakkel, G. (2010) A comparison of two validated tests for upper limb function after stroke: The Wolf Motor Function Test and the Action Research Arm Test. Journal of Rehabilitation Medicine, 42, 694-696.
• Platz, T., Pinkowski, C., van Wijck, F., Kim, I.H., di Bella, P., & Johnson, G. (2005). Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer Test, Action Research Arm Test and Box and Block Test: a multicentre study. Clinical Rehabilitation, 19(4), 404-411.
• Rabadi, M.H. & Rabadi, F.M. (2006). Comparison of the action research arm test and the Fugl-Meyer Assessment as measures of upper-extremity motor weakness after stroke. Archives of Physical of Medicine Rehabilitation, 87, 962-966.
• van der Lee, J.H, Beckerman, H., Lankhorst, G.J., Bouter, L.M. (2001a). The responsiveness of the Action Research Arm Test and the Fugl-Meyer Assessment Scale in chronic stroke patients. Journal of Rehabilitation Medicine, 33, 110-113.
• Van der Lee, J.H, Groot, V., Beckerman, H., Wagenaar, R.C., Lankhorst, G.J., Bouter, L.M. (2001b). The intra-rater and interrater reliability of the action research arm test: a practical test of upper extremity function in patients with stroke. Archives of Physical of Medicine Rehabilitation, 82, 14-19.
• Van der Lee, J.H, Roorda, L.D., & Lankhorst, G.J. (2002). Improving the Action Research Arm Test: a unidimensional hierarchical scale. Clinical Rehabilitation, 16, 646-653.
• Yozbatiran, N., Der-Yerghiaian, L., & Cramer, S.C. (2008). A standardized approach to performing the action research arm test. Neurorehabilitation & Neural Repair, 22(1), 78-90.
• Wagenaar, R.C., Meijer, O.G., van Wieringen, P.C., Kuik, D.J., Hazenberg, G.J., Lindeboom, J., et al. (1990). The functional recovery of stroke: a comparison between neuro-developmental treatment and the Brunnstrom method. Scandinavian Journal of Rehabilitation and Medicine, 22, 1-8.

See the measure

How to obtain the Action Research Arm Test:

The ARAT can be obtained in the study by Lyle (1981), Hsieh et al. (1998), Van der Lee et al. (2002), Rabadi & Rabadi (2006), and Yozbatiran et al. (2008) and from the website: http://www.aratest.eu/Index_english.htm Standardized equipment can be purchased from the website: http://www.aratest.eu/ or from http://www.saliarehab.com/.

Purpose

The Box and Block Test (BBT) measures unilateral gross manual dexterity. It is a quick, simple and inexpensive test. It can be used with a wide range of populations, including clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

In-Depth Review

Purpose of the measure

The Box and Block Test (BBT) measures unilateral gross manual dexterity. It is a quick, simple and inexpensive test. It can be used with a wide range of populations, including clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Available versions

The original version of the BBT was developed, in 1957, by Jean Hyres and Patricia Buhler. This version was modified into the current one by E. Fuchs and P. Buhler (Cromwell, 1976). In 1985, normative data on the BBT was established by Mathiowetz, Volland, Kashman, and Weber.

Features of the measure

Items:

The BBT is composed of a wooden box divided in two compartments by a partition and 150 blocks. The BBT administration consists of asking the client to move, one by one, the maximum number of blocks from one compartment of a box to another of equal size, within 60 seconds. The box should be oriented lengthwise and placed at the client’s midline, with the compartment holding the blocks oriented towards the hand being tested. In order to practice and register baseline scores, the test should begin with the unaffected upper limb. Additionally, a 15-second trial period is permitted at the beginning of each side. Before the trial, after the standardized instructions are given to clients, they should be advised that their fingertips must cross the partition when transferring the blocks, and that they do not need to pick up the blocks that might fall outside of the box (Mathiowetz, Volland, Kashman, & Weber, 1985-1).

Scoring:

Clients are scored based on the number of blocks transferred from one compartment to the other compartment in 60 seconds (Mathiowetz et al., 1985-1). Higher scores are indicative of better manual dexterity. During the performance of the BBT, the evaluator should be aware of whether the client’s fingertips are crossing the partition. Blocks should be counted only when this condition is respected. Furthermore, if two blocks are transferred at once, only one block will be counted. Blocks that fall outside the box, after trespassing the partition, even if they don’t make it to the other compartment, should be counted.

Mathiowetz et al. (1985-1) reported that healthy male adults, aged 20 to 80 years, transfer an average of 77 blocks (SD ±11.6) with the right hand and 75 blocks (SD ±11.4) with the left hand within the 60 second limit. Scores for normal healthy men, aged 60 years old or more ranged from 61 to 70 blocks. Healthy female adults, aged 20 to 80 years, transfer an average of 78 blocks (SD ±10.4) with the right hand and 76 blocks (SD ±9.5) with the left hand. Scores for normal healthy women, aged 60 years old or more, ranged from 63 to 76 blocks. The score on the BBT and age are inversely correlated, meaning that average scores on the BBT decrease with older age.

Time:

The BBT requires 2 to 5 minutes to administer (Finch, Brooks, Stratford, & Mayo, 2002; Mathiowetz et al., 1985-1).

Subscales:

None.

Equipment:

The standardized equipment consists of:
A wooden box dimensioned in 53.7 cm x 25.4 cm x 8.5 cm. The partition should be placed at the middle of the box, dividing it in two containers of 25.4 cm each. (Mathiowetz et al., 1985-1).
150 wooden cubes – 2.5 cm in size (Mathiowetz et al., 1985-1). Stopwatch.

Training of administrator:

None typically reported.

Alternative forms of the Box and Block Test

None.

Client suitability

Can be used with:

• Clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Should not be used in:

• The BBT cannot be used with clients who have severe upper extremity impairment.
• The BBT cannot be used with clients with severe cognitive impairment.

In what languages is the measure available?

There are no official translations of the BBT. The specific instructions provided to the client are in English. Clinicians and researchers may be using “home-grown” translations of the instructions as evidenced from peer-reviewed publication from Sweden, French Canada, Italy and Germany that have used the BBT as an outcome measure. (Broeren, Rydmark, Bjorkdahl, & Sunnerhagen, 2007; Dannenbaun, Michalsen, Desrosiers, & Levin, 2002; Mercier & Bourbonnais, 2004; Platz, Pinkowski, Kim, di Bella, & Johnson, 2005; Schneider, Schonle, Altenmuller, & Munte, 2007).

Summary

What does the tool measure?	Unilateral gross manual dexterity.
What types of clients can the tool be used for?	The BBT can be used with, but is not limited to clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
Time to administer	From 2 to 5 minutes.
Versions	There are no alternative versions.
Other Languages	There are no official translations.
Measurement Properties
ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: No studies have examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the BBT. Test-retest: Two studies have examined the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of the BBT. Both reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). using ICC’s. Inter-rater: Two studies have examined the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the BBT and reported excellent inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. using correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. coefficients and ICC. One study used Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. and the other, ICC and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. .
ValidityThe degree to which an assessment measures what it is supposed to measure.	Criterion: Concurrent: One study has examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.” of the BBT and reported adequate to excellent correlations with the Action Research Arm Test (ARAT) and the Nine-Hole Peg Test (NHPT) at pre and post-treatment. Predictive: One study has examined predictive validityA form of criterion validity that examines a measure’s ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. and reported that the BBT, compared to the NHPT, the Frenchay Arm Test, Grip Strength and the StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Rehabilitation Assessment of Movement (STREAM) was the best predictor of upper limb function 5 weeks post-stroke. Construct: Convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other. : Three studies have examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other. of the BBT and reported excellent correlations between the BBT and the Minnesota Rate of Manipulation Test, the ARAT, the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale and the motor function score of the Fugl-Meyer Assessment (FMA). Adequate correlations were reported between the BBT and the SMAF, the Ashworth scale and the Passive Joint Motion/Joint Pain subscore of the FMA. Poor correlations were reported between the BBT and the Sensation subscore of the FMA and the Modified Barthel Index.
Floor/Ceiling Effects	No studies have examined floor/ceiling effects of the BBT
SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also "Specificity." / SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).	No studies have examined sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” /specificity of the BBT
Does the tool detect change in patients?	Two studies have examined the responsivenessThe ability of an instrument to detect clinically important change over time. of the BBT and reported that the BBT has moderate to large Standardized Response MeanThe standardized response mean (SRM) is calculated by dividing the mean change by the standard deviation of the change scores. , therefore, is able to detect change in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..
Acceptability	The BBT should not be used clients with severe upper extremity impairment and severe cognitive impairments.
Feasibility	The administration of the BBT is quick and simple, however requires standardized equipment.
How to obtain the tool?	The BBT instructions can be obtained in the study by Mathiowetz et al. (1985) Standardized equipment can be obtained at the website: http://www.sammonspreston.com/Supply/Product.asp?Leaf_Id=7531

Psychometric Properties

Overview

We conducted a literature search to identify all relevant publications on the psychometric properties of the Box and Block Test (BBT) in healthy individuals and individuals with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. We identified four studies. The BBT appears to be responsive in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Floor/Ceiling Effects

No studies have examined floor/ceiling effects of the BBT.

Reliability

Test-retest:
Desrosiers, Bravo, Hebert, Dutil, and Mercier (1994) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the BBT in 34 elderly with upper limb sensorimotor impairments from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=13) and other conditions. Participants were re-assessed with a 1-week interval by the same rater and under the same conditions. The test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
for the BBT was reported as excellent (ICC = 0.97; ICC = 0.96) for the right and left hand, respectively.

Platz, Pinkowski, van Wijck, Kim, di Bella, and Johnson (2005) estimated test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the BBT, the Action Research Arm Test (Lyle, 1981), and the Fugl-Meyer Assessment (FMA) upper extremity items including items from the motor function, sensation and passive joint motion/joint pain sub-scores, (Fugl-Meyer, Jääskö, Leyman, Olsson, & Steglind, 1975) in 23 participants with upper extremity paresis either from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, or traumatic brain injury. The participant’s most affected arm was re-assessed after a 1-week interval by the same rater. The test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the BBT, as calculated using ICC’s and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.96 and r = 0.97).
Note: This result applies only to the most affected upper limb.

Inter-rater:
Mathiowetz, Volland, Kashman, and Weber (1985-1) assessed the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the BBT in 26 healthy young females. Participants were evaluated simultaneously and independently by two raters. Pearson correlationcoefficients showed excellent agreement (r = 1.00; r = 0.99) for the right and left hand, respectively.
Note: Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient is not the statistical analysis of choice for assessing inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
as it may artificially inflate agreement.

Platz et al. (2005) as described earlier also analyzed inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the BBT, the Action Research Arm Test (Lyle, 1981), and the FMA upper extremity items including items from the motor function, sensation and passive joint motion/joint pain sub-scores (Fugl-Meyer et al., 1975) in 44 individuals with upper limb paresis either from strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., multiple sclerosis, or traumatic brain injury. Participants had the most affected arm videotaped and scored independently by two raters. Inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
for the BBT, as calculated using the ICC and Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was excellent (ICC = 0.99 and r = 0.99).
Note: This result applies only to the most affected upper limb.

Validity

Content:

Not available.

Criterion:

Concurrent:
No gold standardA measurement that is widely accepted as being the best available to measure a construct.
exists against which to compare the BBT.

Lin, Chuang, Wu, Hsieh and Chang (2010) compared the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the BBT, Action Research Arm Test (ARAT) and Nine-Hole Peg Test (NHPT) for evaluating hand dexterity in 59 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. The Fugl-Meyer Assessment (FMA), Motor Activity Log (MAL) and StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale (SIS) were also administered to assess the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the BBT, ARAT and NHPT. Using Spearman rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient, the BBT, ARAT and NHPT were found to have adequate to excellent correlations at pre-treatment (ranging from rho=-0.55 to -0.80) and post-treatment (ranging from rho=-0.57 to -0.71). In addition, the BBT and ARAT were found to have adequate correlations with the FMA, MAL and SIS (ranging from rho=0.31-59); however, the NHPT had only poor to adequate correlations with the FMA and MAL (ranging from rho=-0.16 to -0.33); and adequate to excellent correlations with the SIS (ranging from rho=-0.58 to -0.66). When considering both the results of responsivenessThe ability of an instrument to detect clinically important change over time.
and validation components of the study, the BBT and ARAT are believed to be more appropriate than the NHPT for evaluating dexterity.

Predictive:
Higgins, Mayo, Desrosiers, Salbach and Ahmed (2005) estimated wheter the BBT, Nine-Hole Peg Test (Kellor, Frost, Silberberg, Iversen, & Cummings, 1971; Mathiowetz, Weber, Kashman, & Volland, 1985-2), Frenchay Arm Test (Heller, Wade, Wood, Sunderland, Hewer, & Ward, 1987), Grip Strength (Mathiowetz, Kashman, Volland, Weber, Dowe, & Rogers, 1985-3), and StrokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. Rehabilitation Assessment of Movement (STREAM – Daley, Mayo, Wood-Dauphine, Danys, & Cabot, 1997) were able to predict upper limb function, measured by the BBT, at 5 weeks post-stroke. Predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the BBT was measured in 55 participants with acute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Assessments were performed at two points in time: one and five weeks post-stroke. Compared to the other upper limb performance tests, the BBT when performed at one week post-stroke, was the best predictor of upper limb function at five months post-stroke, followed by the STREAM.

Construct:

Convergent/Discriminant:
Cromwell (1976) examined the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the BBT by comparing it to the Minnesota Rate of Manipulation Test (American Guidance Service, 1969) in an unspecified population. The correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between BBT and the Minnesota Rate of Manipulation Test was excellent (r = 0.91).

Desrosiers et al. (1994) assessed the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the BBT by comparing it to the Functional Autonomy Measurement System – FAMS, known as the SMAF in French (Hebert, Carries, & Bilodeau, 1988), and to the Action Research Arm Test (ARAT – Lyle, 1981) in 104 elderly with upper limb impairments secondary to strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=53) amongst other conditions. Excellent correlations (r = 0.80) were found between the BBT and the ARAT. Adequate pearson correlations were found between the BBT and the FAMS (r = 0.47; r = 0.51) for the right and left hand, respectively.

Platz et al. (2005) tested the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the BBT by comparing it to the Action Research Arm Test (ARAT – Lyle, 1981) and to the Fugl-Meyer Assessment (FMA)upper extremity items including items from the motor function, sensation and passive joint motion/joint pain sub-scores (Fugl-Meyer et al., 1975) using Spearman CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, in 56 participants with upper extremity paresis either from strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=37) or other conditions. Excellent correlations were found between the BBT and the ARAT (r = 0.95) and the Motor Function sub-score (r = 0.92) of the FMA. Furthermore, the BBT was correlated with more general measures of impairment and activity limitation, such as the Ashworth Scale (Ashworth, 1964), the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale (Adams, Meador, Sethi, Grotta, & Thomson, 1986) and the Modified Barthel Index (Collin, Wade, Davies, & Horne, 1988). Excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
was found between the BBT and the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale (r = -0.67). Adequate correlations were found between the BBT and the passive joint motion/joint pain sub-score of the FMA (r = 0.43) and the Ashworth Scale (r = -0.38). Poor correlations were found between the BBT and the sensation sub-score of the FMA (r = 0.28) and the Modified Barthel Index (r = 0.04).
Note: Negative correlations are observed because a high score on the BBT indicates better performance, whereas a low score on the Hemispheric StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Scale or the Ashworth Scale indicates better performance.

Known groups:
No studies have examined known groups validityKnown groups validity is a form of construct validation in which the validity is determined by the degree to which an instrument can demonstate different scores for groups know to vary on the variables being measured.
of the BBT.

Responsiveness

Higgings et al. (2005) evaluated the responsivenessThe ability of an instrument to detect clinically important change over time.
on the BBT, Frenchay Arm Test (Heller et al., 1987), Grip strength (Mathiowetz et al., 1985-3) and the StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Rehabilitation Assessment of Movement (STREAM – Daley et al., 1997) in 50 participants with acute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were assessed one and four weeks post-stroke. The Standardized Response MeanThe standardized response mean (SRM) is calculated by dividing the mean change by the standard deviation of the change scores.
(SRM) was used to calculate responsivenessThe ability of an instrument to detect clinically important change over time.
. Amongst these upper extremity performance tests, the BBT was the most sensitive to detecting change, having a large SRM of 0.8.
Note: SRM is a variant of effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
and higher values indicate better responsivenessThe ability of an instrument to detect clinically important change over time.
.

Lin, Chuang, Wu, Hsieh and Chang (2010) evaluated the responsivenessThe ability of an instrument to detect clinically important change over time.
of the BBT, the Action Research Arm Test (ARAT) and the Nine-Hole Peg Test (NHPT) for evaluating hand dexterity in 59 patients with subacute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (< 6-months) and Brunnstrom stage IV to VI for proximal and distal upper extremity function. Patients were randomly assigned to receive constraint-induced therapyA form of intervention that involves restraining the unaffected upper or lower extremity in order to encourage movement of the affected limbs. For persons with USN, constraint-induced therapy involves restraining the unaffected arm or hand using a sling or padded mitt, in order to promote visual scanning and movement in the neglected hemispace.
, bilateral arm training or control treatment and received 2 hours of therapy, 5 days per week for 3 weeks. Assessments were performed at baseline and 3 weeks. Using Standardized Response MeanThe standardized response mean (SRM) is calculated by dividing the mean change by the standard deviation of the change scores.
(SRM) to calculate responsivenessThe ability of an instrument to detect clinically important change over time.
, the BBT, ARAT and NHPT were all found to have moderate SRM (0.74, 0.64, 0.79 respectively), indicating sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
for detecting change in hand dexterity. When considering both the results of responsivenessThe ability of an instrument to detect clinically important change over time.
and validation components of the study, the BBT and ARAT are believed to be more appropriate than the NHPT for evaluating dexterity.

References

• American Guidance Service. The Minnesota Rate Manipulative Tests. Examiner’s manual. Circle Pines, (MN): Author; 1969.
• Adams, R.J., Meador, K.J., Sethi, K.D., Grotta, J.C., & Thomson, D.S. (1986). Graded neurologic scale for the use in acute hemispheric stroke treatment protocols. Stroke 18, 665-669.
• Ashworth, B. (1964). Preliminary trial of carisoprodol in multiple sclerosis. Practitioner, 192, 540-542.
• Broeren, J., Rydmark, M., Bjorkdahl, A., & Sunnerhagen, K.S. (2007). Assessment and training in a 3-dimensional virtual environment with haptics: a report on 5 cases of motor rehabilitation in the chronic stage after stroke. Neurorehabilitation & Neural Repair, 21(2), 180-189.
• Collin, C., Wade, D.T., Davies, S., & Horne, V. (1988). The Barthel ADL Index: a reliability study. International Disability Study, 10, 61-63.
• Cromwell, F.S (1965). Occupational therapists manual for basic skills assessment: primary prevocational evaluation. Pasadena, (CA): Fair Oaks Printing; 29-31.
• Daley, K., Mayo, N.E., Wood-Dauphinee, S., Danys, I., & Cabot, R. (1997). Verification of the Stroke Rehabilitation Assessment of Movement (STREAM). Physiotherapy Canada, 49, 269-278.
• Dannenbaum, R.M., Michaelsen, S.M., Desrosiers, J., & Levin, M.F. (2002). Development and validation of two new sensory tests of the hand for patients with stroke. Clinical Rehabilitation, 16(6), 630-639.
• Desrosiers, J., Bravo, G., Hébert, R., Dutil, É., & Mercier, L. (1994). Validation of the box and block test as a measure of dexterity of elderly people: reliability, validity and norms studies. Archives of Physical Medicine and Rehabilitation, 75, 751-755.
• Desrosiers, J., Rochette, A., Hebert, R., & Bravo, G. (1997). The Minnesota manual dexterity test: reliability, validity and reference values studies with healthy elderly People. Canadian Journal of Occupational Therapy, 64(5), 270-276.
• Finch, E., Brooks, D., Stratford,P.W, & Mayo, N.E. (2002). Physical Outcome Measures: A guide to enhance physical outcome measures. Ontario, Canada: Lippincott, Williams & Wilkins.
• Fugl-Meyer, A.R., Jääskö, L., Leyman, I., Olsson, S., & Steglind, S. (1975). The post-stroke hemiplegic patient 1. A method for evaluation of physical performance. Scandinavian Journal of Rehabilitation Medicine, 7, 13-31.
• Hébert, R., Carrier, R., & Bilodeau, A. (1988). The functional autonomy measurement system (SMAF): description and validation of an instrument for the measurement of handicaps. Age Ageing, 17, 293-302.
• Heller, A., Wade, D.T., Wood, V.A., Sunderland, A., Hewer, R., & Ward, E. (1987). Arm function after stroke: measurement and recovery over the first three months. Journal of Neurology, Neurosurgery & Psychiatry, 50(6), 714- 719.
• Higgins, J., Mayo, N.E., Desrosiers, J., Salbach, N.M., & Ahmed, S. (2005). Upper-limb function and recovery in the acute phase poststroke. Journal of Rehabilitation Research & Development, 42(1), 65-76.
• Jebsen, R.H., Taylor, N., Trieschmann, R.B., Trotter, M.J., & Howard, L.A. (1969). An objective and standardized test of hand function. Archives of Physical Medicine and Rehabilitation, 50, 311-319.
• Kellor, M., Frost, J., Silberberg, N., Iversen, I., & Cummings R. (1971). Hand strength and dexterity. American Journal of Occupational Therapy, 25, 77-83.
• Lin, K-C., Chuang, L-L., Wu, C-Y., Hseih, Y-W. & Chang, W-Y. (2010). Responsiveness and validity of three dexterous function measures in stroke rehabilitation. Journal of Rehabilitation Research and Development, 47(6), 563-572.
• Lyle, R.C. (1981). A performance test for assessment of upper limb function in physical rehabilitation treatment and research. International Journal of Rehabilitation and Research, 4, 483-492.
• Mathiowetz, V., Volland, G., Kashman, N., & Weber, K. (1985-1). Adult norms for the box and block test of manual dexterity. American Journal of Occupational Therapy, 39, 386-391.
• Mathiowetz, V., Weber, K., Kashman, N., & Volland, G. (1985-2). Adult norms for the nine hole peg test of finger dexterity. Occupational Therapy Journal of Research, 5, 24 -33.
• Mathiowetz, V., Kashman, N., Volland, G., Weber, K., Dowe, M., & Rogers, S. (1985-3). Grip and pinch strength: normative data for adults. Archives of Physical and Medicine and Rehabilitation, 66, 69-72.
• Mercier, C. & Bourbonnais, D. (2004). Relative shoulder flexor and handgrip strength is related to upper limb function after stroke. Clinical Rehabilitation, 18(2), 215-221.
• Platz, T., Pinkowski, C., van Wijck, F., Kim, I.H., di Bella, P., & Johnson, G. (2005). Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer Test, Action Research Arm Test and Box and Block Test: a multicentre study. Clinical Rehabilitation, 19(4), 404-411.
• Schneider, S., Schonle, P.W., Altenmuller, E., & Munte, T.F. Using musical instruments to improve motor skill recovery following a stroke. Journal of Neurology, 254(10), 1339-1346.
• Tiffin, J. (1968). Purdue Pegboard Examiner Manual. Chicago, USA: Science Research Associates.

See the measure

How to obtain the BBT

The BBT instructions can be obtained in the study by Mathiowetz et al. (1985)

Standardized equipment can be obtained at the website:
http://www.sammonspreston.com/Supply/Product.asp?Leaf_Id=7531

By clicking here, you can access a video showing how to administer the assessment.

Purpose

The Chedoke Arm and Hand Activity Inventory (CAHAI) is a functional assessment of the recovering arm and hand after strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. The CAHAI compliments the Chedoke-McMaster Stroke Assessment (Barreca, Stratford, Masters, Lambert, Griffiths, and McBay, 2006).

In-Depth Review

Purpose of the measure

The Chedoke Arm and Hand Activity Inventory (CAHAI) is a functional assessment of the recovering arm and hand after strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. The CAHAI compliments the Chedoke-McMaster Stroke Assessment (Barreca, Stratford, Masters, Lambert, Griffiths, and McBay, 2006).

Available versions

The CAHAI was developed by Barreca, Gowland, Stratford, Huijbregts, Griffiths, Torresin, Dunkley, Miller, and Masters in 2004 to address the need for a valid, clinically relevant, and responsive functional assessment of the recovering paretic upper limb.

Three shortened versions of the CAHAI were developed by Barreca, Stratford, Masters, Lambert, Griffiths, and McBay in 2006. The shortened versions have 7, 8 or 9 items and are identified as CAHAI-7, CAHAI-8, CAHAI-9, respectively.

Features of the measure

Items:

The original CAHAI consists of 13 functional items that are non-gender specific, involve both upper limbs, and incorporates a range of movements and grasps that reflect stages of motor recovery following strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. The following items were generated from a review of the scientific literature on strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., as well as from input from individuals with stroke and their families (Barreca et al., 2004):

1. Open a jar of coffee
2. Dial 911
3. Draw a line with a ruler
4. Pour a glass of water
5. Wring out a washcloth
6. Do up five buttons
7. Dry back with a towel
8. Put toothpaste on a toothbrush
9. Cut medium consistency putty
10. Clean eye glasses
11. Zip up a zipper
12. Place a container on a table
13. Carry a bag up the stairs

The CAHAI-7 utilizes the first 7 items, CAHAI-8 the first 8 items, and CAHAI-9 the first 9 items. The 13 items together represent the original CAHAI (Barreca et al., 2006). On average, clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. consider items 1, 2, 4 and 12 easy to perform; items 8, 10, 11, and 13 moderately difficult; and items 3, 6, 7, and 9 the most difficult (Barreca et al., 2004).

Detailed administration guidelines are in the development manual that can be obtained can be obtained by visiting the official website: http://www.cahai.ca

Scoring:

Each item of the CAHAI is scored on a 7-point quantitative scale, similar to the scale used in the Functional Independence Measure (FIM) (Keith, Granger, Hamilton, & Sherwin, 1987)

A score of

• 1 = client needs total assistance and the weak upper limb performs less than 25% of the task;
• 2 = client needs maximal assistance and the weak upper limb performs 25% to 49% of the task. There are no signs of arm or hand manipulation, only stabilization;
• 3 = client needs moderate assistance and the weak upper limb performs 50% to 74% of the task. Begins to show signs of arm or hand manipulation;
• 4 = client needs minimal assistance (light touch) and the weak upper limb performs more than 75% of the task;
• 5 = client requires supervision, coaxing, or cueing;
• 6 = client requires use of assistive devicesAssistive devices are any piece of equipment that you use to make your daily activities easier to perform.
  or requires more than reasonable time, or there are safety concerns; and
• 7 = total independence in completing the task.

The minimal possible score for the CAHAI is 13 and the maximum is 91, with higher scores indicating greater functional independence (Barreca et al., 2004; Barreca, Stratford, Lambert, Masters, & Streiner, 2005; Barreca, Stratford, Masters, Lambert, & Griffiths, 2006b).

The affected limb is also scored according to its positioning and functioning during test performance. The therapist should record the performance of the affected limb on each item by checking the appropriate box. The scoring table for the CAHAI is as follows: (Barreca et al., 2004):

Items	Affected Limb
1) Open a jar of coffee	Holds jar	Holds lid
2) Call 911	Holds receiver	Dials phone
3) Draw a line with ruler	Holds ruler	Holds pen
4) Put toothpaste on toothbrush	Holds toothpaste	Holds brush
5) Cut medium consistency putty	Holds knife	Holds fork
6) Pour a glass of water	Holds glass	Holds pitcher
7) Clean a pair of eyeglasses	Holds glasses	Wipes lenses
8) Zip up the zipper	Holds zipper	Holds zipper pull
9) Dry back with towel	Reaches for towel	Grasps towel end

Note: Standardized instructions on scoring can be obtained by visiting the official website: http://www.cahai.ca

Time:

The time to administer and score the CAHAI is approximately 25 minutes (Barreca et al., 2004; Barreca et al., 2006).

Subscales:

None

Equipment required:

CAHAI-7

Version (Items 1-7) requires all items in Equipment List A

Equipment List A

• height adjustable table
• chair/wheelchair without armrests
• dycem
• 200g jar of coffee
• push-button telephone
• 12″/30cm ruler
• 8.5″ x 11″ paper
• pencil
• 2.3L plastic pitcher with lid filled with 1600 ml. Water
• 250 ml plastic cup
• wash cloth
• wash basin (24.5 cm. in diameter, height 8 cm.)
• Pull-on vest with 5 buttons (one side male & one side female), buttons (1.5 cm. In diameter, 7 cm. apart)
• bath towel (65cm X 100cm)

CAHAI-8

Version (Items 1-8) requires all items in Equipment List A and B

Equipment List B

• 75ml toothpaste with screw lid, >50% full
• toothbrush

CAHAI-9

Version (Items 1-9) requires all items in Equipment List A, B, and C

Equipment List C

• dinner plate (Melamine or heavy plastic, 25 cm. in diameter)
• medium resistance putty
• knife and fork
• built up handles the length of the utensil handle

CAHAI-13

Version (Items 1-13) requires all items in Equipment List A, B, C, and D

Equipment List D

• 27″/67cm metal zipper in polar fleece poncho
• eyeglasses
• handkerchief
• Rubbermaid 38L container (50 x 37 x 27cm)
• 4 standard size steps with rail
• plastic grocery bag holding 4lb/2kg weight

Training:

Training may be provided by the authors as a half-day workshop. There is a training DVD available in English for a cost of $29.00 Canadian including shipping. Only cheque or money orders are processed.

Alternative forms of the CAHAI

CAHAI-7, CAHAI-8, CAHAI-9

Client suitability

Can be used with:

• Clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Should not be used in:

• To date, there is no information on restrictions of using the CAHAI.

In what languages is the measure available?

English, French, German, Hebrew, Italian

Summary

What does the tool measure?	The CAHAI assess upper limb functional recovery.
What types of clients can the tool be used for?	The CAHAI can be used with, but is not limited to clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
Time to administer	An average of 20 to 25 minutes
Versions	CAHAI, CAHAI-9, CAHAI-8, CAHAI-7.
Other Languages	English, French, German, Hebrew and Italian.
Measurement Properties
ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: Two studies have examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the CAHAI and its shortened versions and reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. using Cronbach’s alpha. Test-retest: One study examined the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of the CAHAI and reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). using using the Intraclass Correlation Coefficient (ICC)Intraclass correlation (ICC) is used to measure inter-rater reliability for two or more raters. It may also be used to assess test-retest reliability. ICC may be conceptualized as the ratio of between-groups variance to total variance.. Intra-rater: No studies have examined the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the CAHAI. Inter-rater: One study examined the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the CAHAI and reported excellent inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. using ICC.
ValidityThe degree to which an assessment measures what it is supposed to measure.	Content: One study examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension. of the CAHAI and reported that items were generated from a review of scientific literature and from input from clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., their family and caregivers. Items with poor frequency endorsement, difficulty to be standardized, and high inter-item correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. were eliminated. Criterion: Concurrent: One study examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.” of the CAHAI and the CAHAI-9 and reported that the CAHAI-9 was not able to predict individual scores and individual change scores of the CAHAI, using regression analysis. Predictive: No studies have examined the predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. of the CAHAI. Construct: Convergent: Three studies examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other. of the CAHAI and reported excellent correlations between all versions of the CAHAI and the Action Research Arm Test, and all versions of the CAHAI and the Chedoke-McMaster Stroke Assessment (CMSA), and poor to moderate correlations between the CAHAI and the CMSA shoulder pain score, using Pearson CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. . Known Groups: Three studies examined longitudinal/known groups validityThe degree to which an assessment measures what it is supposed to measure. of all versions of the CAHAI and reported that all versions are able to distinguish changes between subjects with acute and chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., and mild from severe impairments, using ROC curve (Receiver Operation Characteristic).
Floor/Ceiling Effects	No studies have examined the floor/ceiling effects of the CAHAI.
SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” / SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).	No studies have examined the sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” /specificity of the CAHAI.
Does the tool detect change in patients?	One study examined the responsivenessThe ability of an instrument to detect clinically important change over time. of the CAHAI and reported that the minimal detectable change between two evaluations in stable patients was 6.3 points.
Acceptability	The CAHAI is highly accepted by clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. since is made up of real-life and non-gender specific items.
Feasibility	The administration of the CAHAI is easy and quick to perform.
How to obtain the tool?	The CAHAI can be obtained free of charge by visiting the official website: http://www.cahai.ca

Psychometric Properties

Overview

We conducted a literature search to identify all relevant publications on the psychometric properties of the Chedoke Arm and Hand Activity Inventory (CAHAI) in individuals with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. We identified four studies. The CAHAI appears to be responsive in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Floor/Ceiling Effects

No studies have examined floor/ceiling effects of the CAHAI.

Reliability

Internal ConsistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.:
Barreca, Gowland, Stratford, Huijbregts, Griffiths, Torresin, Dunkley, Miller, and Masters (2004) assessed the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the CAHAI in 100 clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the CAHAI, as calculated using Cronbach’s Coefficient Alpha was excellent (α = 0.98).

Barreca, Stratford, Masters, Lambert, Griffiths, and McBay (2006) examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the CAHAI-7, CAHAI-8, and CAHAI-9 in 39 clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of all shortened versions of the CAHAI, as calculated using Cronbach’s Coefficient Alpha, was excellent (α = 0.97; α = 0.98; α = 0.98, respectively).

Test-retest:
Barreca et al. (2006) examined the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the shortened version of the CAHAI in 39 clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were stratified into two different groups based on the amount of expected improvement. Participants were re-assessed following a 36 hour interval. The test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
as calculated using Intraclass Correlation Coefficient (ICC)Intraclass correlation (ICC) is used to measure inter-rater reliability for two or more raters. It may also be used to assess test-retest reliability. ICC may be conceptualized as the ratio of between-groups variance to total variance. was excellent for all shortened versions: CAHAI-7 (ICC = 0.96), CAHAI-8 (ICC = 0.97), and CAHAI 9 (ICC = 0.97).

Intra-rater:
No studies have examined the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
of the CAHAI.

Inter-rater:
Barreca, Stratford, Lambert, Masters, and Streiner (2005) assessed the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the CAHAI in 39 clients with stroke. Participants were stratified into two different groups based on the amount of expected improvement. Participants were re-assessed following a 36 hours interval. The inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
as calculated using Intraclass Correlation Coefficient (ICC)Intraclass correlation (ICC) is used to measure inter-rater reliability for two or more raters. It may also be used to assess test-retest reliability. ICC may be conceptualized as the ratio of between-groups variance to total variance., was excellent (ICC = 0.98).

Validity

Content:

Barreca et al. (2004) performed a literature review to generate items for the CAHAI. From this review, 177 items were selected. Eighty-one clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., their families and caregivers were surveyed about important and relevant items regarding strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. recovery, which generated an additional 574 items. To reduce the 725 generated items to 26 items, only bilateral, gender-neutral items, that fell into the domains identified by the clients as important that were easy to obtain were kept. This version, with 26 items, was then tested in 20 participants with stroke. Items that were difficult to standardize or those with the potential for safety concerns were eliminated. Items with a high degree of difficulty were added in order to minimize possible ceiling effects. Inter-item correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
analyses of this new version (which contained 25 items), identified some redundant items (r > 0.90). Items with poor frequency endorsement, difficulty to standardize and high inter-item correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
were eliminated, resulting in the 13 finalized items.

Criterion:

Concurrent:
Barreca, Stratford, Masters, Lambert, & Griffiths (2006b) examined the ability of the CAHAI-9 to predict the scores and change scores of the original CAHAI in 105 clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Mean scores and mean change scores of the CAHAI-9 accurately predicted means scores and mean change scores of the CAHAI. However, individual scores and individual change scores of the CAHAI-9 displayed moderate variability in predicting individual scores and change scores of CAHAI. The findings indicate that the CAHAI-9 should not be administered with the intent to predict the CAHAI.

Predictive:
No studies have examined the predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the CAHAI.

Construct:

Convergent/Discriminant:
Barreca et al. (2005) estimated convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the CAHAI by comparing it to Chedoke-McMaster StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Assessment (CMSA – Gowland, Stratford, Ward, Moreland Torresin, VanHullenaar et al., 1993; Gowland, VanHullenaar, Torresin, et al., 1995) arm-hand sum score, and with the Action Research Arm Test (ARAT – Lyle, 1981) in 39 participants with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Assessments were performed at baseline and 2 to 6 weeks later. Correlations, as calculated using Pearson CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
Coefficient were excellent between the CAHAI and the ARAT (r = 0.93) and between the CAHAI and the CMSA arm-hand at baseline (r = 0.81) and at follow up (r = 0.89). In the same study, the authors analyzed discriminant validityMeasures that should not be related are not. Discriminant validity examines the extent to which a measure correlates with measures of attributes that are different from the attribute the measure is intended to assess.
of the CAHAI by comparing it to the CMSA shoulder pain score in the same 39 participants with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. The correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the CAHAI and CMSA shoulder pain score as calculated using Pearson CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
, was adequate at baseline (r = 0.47) and at follow-up (r = 0.39).

Barreca et al. (2006) assessed the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the CAHAI-7, CAHAI-8 and CAHAI-9 by comparing them to the Action Research Arm Test (ARAT), CAHAI and CMSA in 39 individuals with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Pearson Correlations were used. Correlations between the ARAT and CAHAI-7 (r = 0.95), CAHAI-8 (r = 0.95) and CAHAI-9 (r = 0.94) were all excellent , as well as between the CAHAI and all the shortened versions (r = 0.99), and between the CMSA and CAHAI-7 (r = 0.85), CAHAI-8 (r = 0.84), and CAHAI-9 (r = 0.84).

Barreca et al., (2006b) determined the convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the CAHAI-9 and CAHAI by comparing them to the ARAT (Lyle, 1981) in 105 individuals with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Re-assessments were performed with a 36 hours interval. Pearson CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
Coefficients were excellent between the CAHAI-9 and ARAT at baseline (r = 0.93), and at follow-up (r = 0.95), as well as between the CAHAI at baseline (r = 0.93), and at follow-up (r = 0.95).

Known groups:
Barreca et al. (2005) analyzed the longitudinal validityLongitudinal validity is the extent to which changes on one measure will correlate with changes on another measure.
of the CAHAI in 39 clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. by comparing change scores on the CAHAI with change scores on the arm-hand sum and on the shoulder pain dimensions of the Chedoke-McMaster StrokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. Assessment (CMSA – Gowland et al., 1995) and on the Action Research Arm Test (ARAT – Lyle, 1981). Change scores correlations, as calculated using Pearson CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
Coefficient, was excellent between the CAHAI and the ARAT (r = 0.86), adequate between the CAHAI and the CMSA arm-hand sum (r = 0.52) and poor between the CAHAI and the CMSA shoulder pain (r = -0.24). In a second analysis, Barreca et al. (2005) analyzed whether the CAHAI was more adept then the CMSA and the ARAT at distinguishing change in patients with mild/moderate impairments from patients with severe impairments in 39 clients with stroke. Longitudinal/known groups validityThe degree to which an assessment measures what it is supposed to measure.
, as calculated using Receiver Operating Characteristic (ROC) demonstrated an excellent area under the curve for the CAHAI (ROC = 0.95). The ARAT and CMSA presented an adequate area under the curve (ROC = 0.88; ROC = 0.76), respectively.
Note: ROC curve analysis quantifies a measure’s ability to distinguish between groups as an area under the ROC curve. Greater areas indicate the measure is better at discriminating between individuals in the two groups.

Barreca et al. (2006) assessed the longitudinal validityLongitudinal validity is the extent to which changes on one measure will correlate with changes on another measure.
of the CAHAI and its three shortened versions in 39 participants with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were divided according to stroke’s severity into acute and chronic groups. The CAHAI, CAHAI-7, CAHAI-8, and CAHAI-9 were administered at admission and discharge (2 to 6 weeks after admission) to verify which version was more adept to detecting changes in patients with acute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. from patients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Longitudinal/known groups validityThe degree to which an assessment measures what it is supposed to measure.
, as calculated using Receiver Operating Characteristic (ROC) demonstrated an excellent area under the curve for all versions of the CAHAI as follows: CAHAI (ROC = 0.95); CAHAI -7 (ROC = 0.97); CAHAI-8 (ROC = 0.93), and CAHAI-9 (ROC = 0.94), meaning all versions of CAHAI are equally able to distinguish changes between different groups in strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Barreca et al. (2006b) examined the longitudinal validityLongitudinal validity is the extent to which changes on one measure will correlate with changes on another measure.
of the CAHAI, CAHAI-9 and the ARAT in 105 individuals with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were stratified between mild/moderate impairments and severe impairments, and those with mild/moderate impairments were expected to show greater changes across two repeated measures. The three outcome measures were administered at two points in time to verify which of them were more adept to detecting changes in clients with mild/moderate impairment from clients with severe impairment. Longitudinal/known groups validityThe degree to which an assessment measures what it is supposed to measure.
, as calculated using Receiver Operating Characteristics, were adequate for the ARAT (ROC = 0.72), the CAHAI -9 (ROC = 0.82), and the CAHAI (ROC = 0.86). This ROC analysis indicated that the CAHAI was the best measure to detect change among patients with mild/moderate impairment from patients with severe impairment.

Responsiveness

Barreca et al. (2005) assessed the minimal detectable change of the CAHAI in 39 clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were assessed at two points in time: at admission, and after 2 to 6 weeks. For the CAHAI, the minimal detectable change was 6.3 points, meaning that stable patients displayed random fluctuations of 6.3 CAHAI points or less when assessed on two different occasions.

References

• Barreca, S.R., Gowland, C.K., Stratford, P.W., et al. (2004). Development of the Chedoke Arm and Hand Activity Inventory: Theoretical constructs, item generation, and selection. Topics in Stroke Rehabilitation, 11(4), 31- 42.
• Barreca, S.R., Stratford, P.W., Lambert, C.L., Masters, L.M., & Streiner, D.L. (2005). Test-retest reliability, validity, and sensitivity of the Chedoke Arm and Hand Activity Inventory: a new measure of upper-limb function for survivors of stroke. Archives of Physical Medicine and Rehabilitation, 86, 1616-1622.
• Barreca, S.R., Stratford, P.W., Masters, L.M., Lambert, C.L., Griffiths, J., McBay, C. (2006). Validation of three shortened versions of the Chedoke Arm and Hand Activity Inventory. Physiotherapy Canada, 58, 148-156.
• Barreca, S.R., Stratford, P.W., Masters, L.M., Lambert, C.L., Griffiths, J. (2006b). Comparing two versions of the Chedoke Arm and Hand Activity Inventory with the Action Research Arm Test. Physical Therapy, 86(2), 245-253.
• Gowland, C., Stratford, P., Ward, M., Moreland, J., Torresin, W., VanHullenaar, S. et al.(1993). Measuring physical impairment and disability with the Chedoke-McMaster Stroke Assessment. Stroke, 24,58-63.
• Gowland, C., VanHullenaar, S., Torresin, W., et al. (1995). Chedoke-McMaster Stroke Assessment: development, validation, and administration manual. Hamilton, ON, Canada: School of Rehabilitation Science, McMaster University.
• Heller, A., Wade, D.T., Wood, V.A., Sunderland, A., Hewer, R., & Ward, E. (1987). Arm function after stroke: measurement and recovery over the first three months. Journal of Neurology, Neurosurgery & Psychiatry, 50(6), 714-719.
• Keith, R.A, Granger, C.V., Hamilton, B.B., & Sherwin, F.S. (1987). The Functional Independence Measure: a new tool for rehabilitation. In: Eisenberg, M.G. & Grzesiak, R.C. (Ed.), Advances in clinical rehabilitation (pp. 6-18). New York: Springer Publishing Company.
• Kellor, M., Frost, J., Silberberg, N., Iversen, I., & Cummings R. (1971). Hand strength and dexterity. American Journal of Occupational Therapy, 25, 77-83.
• Lyle, R.C. (1981). A performance test for assessment of upper limb function in physical rehabilitation treatment and research. International Journal of Rehabilitation and Research, 4, 483-492.
• Mathiowetz, V., Kashman, N., Volland, G., Weber, K., Dowe, M., & Rogers, S. (1985). Grip and pinch strength: normative data for adults. Archives of Physical and Medicine and Rehabilitation, 66, 69-72.
• Mathiowetz, V., Weber, K., Kashman, N., & Volland, G. (1985b). Adult norms for the nine hole peg test of finger dexterity. Occupational Therapy Journal of Research, 5, 24 -33.

See the measure

How to obtain the CAHAI

The CAHAI can be obtained free of charge by visiting the official website: http://www.cahai.ca

Purpose

The Disabilities of the Arm, Shoulder and Hand (DASH) is a self-report questionnaire that measures disability and symptoms of upper limb musculoskeletal disorders.

In-Depth Review

Purpose of the measure

The Disabilities of the Arm, Shoulder and Hand (DASH) is a self-report questionnaire that measures physical function and symptoms of the upper limb. The DASH can be used for any joint and any musculoskeletal condition of the upper limb (Hudak et al., 1996; Veehof et al., 2002), which permits comparison across upper limb diagnoses (Atroshi et al., 2000). The DASH is intended for discriminative and evaluative purposes (Schmitt & Di Fabio, 2004).

The DASH demonstrates validityThe degree to which an assessment measures what it is supposed to measure.
and responsivenessThe ability of an instrument to detect clinically important change over time.
in proximal and distal upper limb disorders (Beaton et al., 2001). The DASH demonstrated better clinimetric properties than other shoulder disability questionnaires including the Simply Shoulder Test (SST), American Shoulder and Elbow Surgeons Standardised Shoulder assessment Form (ASES) and the Shoulder Pain and Disability Index (SPADI – Bot et al., 2004).

Available versions

The DASH was developed by the American Academy of Orthopedic Surgeons, the Council of the Musculoskeletal Specialty Societies, and the Institute for Work and Health as a region-specific instrument to measure patients’ perception of disability and symptoms associated with any joint or condition of the upper limb (Hudak et al., 1996; Veehof et al., 2002).

The third edition of the DASH has been recently published to incorporate the latest research and new information regarding cross-cultural use of the measure.

Features of the measure

Items:

The DASH consists of 30 items that measure: (a) physical function (21 items); (b) symptom severity (5 items); and (c) social or role function (4 items).

Ability to do the following activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
:

1. Open a tight or new jar
2. Write
3. Turn a key
4. Prepare a meal
5. Push open a heavy door
6. Place an object on a shelf above your head
7. Do heavy household chores (e.g. wash walls, wash floors)
8. Garden or do yard work
9. Make a bed
10. Carry a shopping bag or briefcase
11. Carry a heavy object (over 5kg)
12. Change a light bulb overhead
13. Wash or blow dry your hair
14. Wash your back
15. Put on a pullover sweater
16. Use a knife to cut food
17. Recreational activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
    that require little effort (e.g. card playing, knitting)
18. Recreational activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
    that require taking some force or impact through the arm, shoulder or hand (e.g. golf, hammering, tennis)
19. Recreational activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
    that require you to move the arm freely (Frisbee, badminton)
20. Managing transportation needs (getting from one place to another0
21. Sexual activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
    
22. Extent to which arm, shoulder or hand problems interfered with normal social activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
    with family, friends, neighbours or groups
23. Extent to which arm, shoulder or hand problems limited work or other regular daily activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
    

Severity of the following symptoms:

1. Arm, shoulder or hand pain
2. Arm, shoulder or hand pain when performing activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
   
3. Tingling
4. Weakness
5. Stiffness
6. Difficulty in sleeping
7. Impact on self-image

The DASH also includes two optional modules regarding work and sports/performing arts that investigate the individual’s difficulty:

1. Using the usual technique for the activity (work; sport/instrument)
2. Performing the activity due to arm, shoulder or hand pain
3. Performing the as well as he/she would like
4. Spending the usual amount of time on the activity

Scoring:

The most recent version of the DASH uses a 5-point Likert scaleLikert scaling is one type of response to items in a questionnaire or tool. For example, Likert scaling would have you rate an item such as “I am satisfied with the care I received” on a scale using a 1-to-5 response scale where:
• 1 = strongly disagree
• 2 = disagree
• 3 = undecided
• 4 = agree
• 5 = strongly agree
You will find various options and scaling methods for the number of response choices (1-to-7, 1-to-9, 0-to-4). Odd-numbered scales usually have a middle value that is labelled Neutral or Undecided. Some tools used forced-choice Likert scaling with an even number of responses and no middle neutral or undecided choice. that rates the individual’s difficulties the preceding week. Lower scores indicate no difficulty, limitations or symptoms whereas higher scores indicate inability to perform tasks or extreme difficulties or symptomatology.

Items 1 – 21	1 = no difficulty 2 = mild difficulty 3 = moderate difficulty 4 = severe difficulty 5 = unable
Item 22	1 = not at all 2 = slightly 3 = moderately 4 = quite a bit 5 = extremely
Item 23	1 = not limited at all 2 = slightly limited 3 = moderately limited 4 = very limited 5 = unable
Items 24 – 28	1 = none 2 = mild 3 = moderate 4 = severe 5 = extreme
Optional work and sports/performing arts modules:	1 = no difficulty 2 = mild difficulty 3 = moderate difficulty 4 = severe difficulty 5 = unable

The DASH total score is calculated as a percentage (0=no disability to 100=maximal disability), using the following calculation:

[(Sum of completed responses ÷ number of completed responses) – 1] x 25

The final score for each optional module is calculated as follows:

[(Sum of completed responses ÷ 4) – 1] x 25

Note: A DASH total score cannot be calculated if more than 3 items have not been answered. Total scores for the additional modules cannot be calculated if there are any missing items.

Where 3 or fewer items have been missed, missing responses are replaced by the mean value of the responses to other items before summing.

Please note that earlier versions of the DASH use a different scoring system.

What to consider before beginning:

A study by Ring et al. (2006) showed a strong correlation between the DASH and measures of depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
(Center for Epidemiologic Studies – DepressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
) and anxiety (Pain Anxiety Symptoms Scale) in a sample of 235 patients with discrete hand problems (e.g. carpal tunnel syndrome, de Quervain tenosynovitis, lateral elbow pain, trigger finger, distal radial fracture). Subsequently, Lozano Calderon et al. (2010) conducted a study with 516 patients requiring hand surgery and adjusted DASH scores for the influence of depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
. This resulted in a significant decrease in the mean and standard deviation of DASH scores, although the decrease in variation was small. There was a high correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between DASH and depression-adjusted DASH scores, indicating no notable benefit to adjusting DASH scores for depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
. Given the high incidence of depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
among patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., consideration of the correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between disability and depressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely "pull themselves together" and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
should be considered when using the DASH.

Time:

The DASH takes approximately 5 minutes to administer with patients with musculoskeletal disorders (Bot et al., 2004). Administration with patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. may require more time and support materials.

Training requirements:

No specific training requirements are specified.

Equipment:

No specific equipment is required.

Alternative Forms of the Measure

The QuickDASH is an 11-item questionnaire that was developed from the DASH using a concept-retention’ approach (Beaton et al., 2005). The QuickDASH is comprised of the following items:

1. Open a tight or new jar
2. Do heavy household chores (e.g. wash walls, wash floors)
3. Carry a shopping bag or briefcase
4. Wash your back
5. Use a knife to cut food
6. Recreational activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
   that require taking some force or impact through the arm, shoulder or hand (e.g. golf, hammering, tennis)
7. Extent to which arm, shoulder or hand problems interfered with normal social activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
   with family, friends, neighbours or groups
8. Extent to which arm, shoulder or hand problems limited work or other regular daily activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
   
9. Arm, shoulder or hand pain
10. Tingling
11. Difficulty in sleeping

The QuickDASH also retains the optional work and sports/performing arts modules (Beaton et al., 2005).

Like the DASH, the QuickDASH uses a 5-point Likert rating scale and the total score is calculated as a percentage (0=no disability – 100=most severe disability). At least 10 of the 11 items must be completed for correct use. The QuickDASH demonstrates similar test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
, validityThe degree to which an assessment measures what it is supposed to measure.
and responsivenessThe ability of an instrument to detect clinically important change over time.
to the DASH and may demonstrate better precision in detecting different degrees of disability than the DASH. Although there is a high correlation between the QuickDASH and the DASH, an exact match between the numeric scores of the two assessments is not guaranteed (Beaton et al., 2005). Due to the smaller number of items, the QuickDASH is considered to be more efficient than the DASH (Beaton et al., 2005; Gummesson et al., 2006). However, the DASH is more suitable than the QuickDASH for use when monitoring"The process of checking the task over time for ‘quality control’ and the adjustment of behavior" (Stuss, 2009, p. 9-10)
arm pain and function over time in individual patients.

Client suitability

Can be used with:

• Individuals with upper limb musculoskeletal impairment.
• Due to limited research regarding patient acceptability, the DASH may be more suitable for patients with mild impairment.

Should not be used with:

• N/A

Languages of the measure

Approved translations have been made in the following languages:

• Afrikaans
• Arabic
• Armenian
• Chinese (Hong Kong)
• Chinese (Taiwan)
• Czech
• Danish
• Dutch
• English (Australia)
• English (Hong Kong)
• English (South Africa)
• Finnish
• French Canadian
• French
• German
• Greek
• Hebrew
• Hungarian
• Italian
• Japanese
• Korean
• Lithuanian
• Malay
• Norwegian
• Persian (Iran)
• Polish
• Portugese (Brazil)
• Portugese (Portugal)
• Romanian
• Russian
• Serbian
• Sinhala (Sri Lanka)
• Spanish (Argentina)
• Spanish (Puerto Rico)
• Spanish (Spain)
• Swedish
• Thai
• Turkish

Translations are also in progress for the following languages:

• Croatian
• Estonian
• Filipino
• Isi-Xhosa
• Latvian
• Malayalam
• Slovak
• Spanish (Chile)
• Spanish (Dominican Republic)
• Ukrainian

Summary

What does the tool measure?	Upper extremity disability and pain.
What types of clients can the tool be used for?	Individuals with musculoskeletal disorders of the upper limb.
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
Time to administer	Five minutes.
Versions	DASH QuickDASH
Other Languages	Afrikaans, Arabic, Armenian, Chinese (Hong Kong), Chinese (Taiwan), Czech, Danish, Dutch, English (Australia), English (Hong Kong), English (South Africa), Finnish, French Canadian, French, German, Greek, Hebrew, Hungarian, Italian, Japanese, Korean, Lithuanian, Malay, Norwegian, Persian (Iran), Polish, Portugese (Brazil), Portugese (Portugal), Romanian, Russian, Serbian, Sinhala (Sri Lanka), Spanish (Argentina), Spanish (Puerto Rico), Spanish (Spain), Swedish, Thai, Turkish.
Measurement Properties
ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: No studies have reported on the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the DASH among patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Test-retest: No studies have reported on the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of the DASH among patients with stroke. Intra-rater: No studies have reported on the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the DASH among patients with stroke. Inter-rater: No studies have reported on the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the DASH among patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
ValidityThe degree to which an assessment measures what it is supposed to measure.	Content: The DASH was developed by item generation (clinical expert input, literature review and patient focus groups) and item reduction (expert review, and psychometric and clinimetric analysis). One study that examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension. of the DASH in a sample of patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. suggested a disordered rating scale structure and item hierarchy that is not suitable for clinical use. Criterion: Concurrent: No studies have reported on the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.” of the DASH among patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Predictive: No studies have reported on the predictive validityA form of criterion validity that examines a measure’s ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. of the DASH among patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Construct: Convergent/Discriminant: One study reported moderate correlations between manual ability and pain. Known Groups: No studies have reported on the known-groups validityThe degree to which an assessment measures what it is supposed to measure. of the DASH among patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..
Floor/Ceiling Effects	No studies have reported on the floor/ceiling effects of the DASH among patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..
Does the tool detect change in patients?	No studies have reported on the responsivenessThe ability of an instrument to detect clinically important change over time. among patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..
Acceptability	The DASH is simple to comprehend, quick to complete and is comprised of real-life, non-gender specific items. Due to limited research regarding patient acceptance, this tool may be more suitable for patients with mild impairment.
Feasibility	The DASH is a versatile measure that can be used for clinical or research purposes. However there is insufficient research regarding use of the DASH with patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. and concerns that without testing, the clinical utility of the DASH remains unknown.
How to obtain the tool?	Visit the DASH website for more information: https://dash.iwh.on.ca/

Psychometric Properties

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the DASH. While numerous studies have been conducted with other patient groups, this review specifically addresses the psychometric properties relevant to patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. At the time of publication there was 1 conference paper but no published studies specific to patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Floor/Ceiling Effects

No studies have reported on the floor/ceiling effects of the DASH in a sample of patients with stroke. The DASH demonstrates no floor or ceiling effects in patients with shoulder and combined shoulder-upper limb problems (Bot et al., 2004).

Reliability

Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.:
No studies have examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the DASH in a sample of patients with stroke, although studies conducted among patient groups with other upper limb conditions indicate excellent reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of "true" variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to "noise" in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
(see: Atroshi et al., 2000; Bot et al., 2004; Veehof et al., 2002). However, this may indicate item redundancy (Beaton et al., 2005).

Test-retest:
No studies have examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the DASH in a sample of patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., although studies conducted among patient groups with other upper limb conditions indicate excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
(see: Atroshi et al., 2000; Bot et al., 2004; Beaton et al., 2001).

Intra-rater:
No studies have examined intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
of the DASH in a sample of patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Inter-rater:
No studies have examined inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the DASH in a sample of patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Validity

Content:

The DASH was developed in two stages of item generation and item reduction. The first stage of item generation involved clinical expert input, review of 13 relevant outcome measurement scales and patient focus groups to identify possible items. The second stage of item reduction involved preliminary item review by three content experts, secondary review by a panel of 15 experts for content/face validity and item importance, and subsequent pre-testing on 20 individuals with upper extremity difficulties. Further item reduction was conducted by psychometric and clinimetric analysis among patients with upper limb conditions, including (i) field-testing in a cross-sectional study of 407 patients with various upper limb problems, and (ii) importance- and difficulty- rating in a second sample of 76 patients. This resulted in the 30-item questionnaire (Hudak et al., 1996; Marx et al., 1999).

Lannin et al. (2010) examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension.
of the DASH in a sample of 157 patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Analysis of the original rating scale revealed a disordered structure; Rasch measurement modeling was used to transform ordinal ratings into a collapsed linear measure, which resulted in conformation to expectations of the model. The study also found that the hierarchy of the original 30 items is not appropriate for clinical use as there are few items suitable for the most disabled patient.

Franchignoni et al. (2010) investigated the dimensionality, rating scale diagnostics and model fit of the DASH (Italian version) on a sample of 238 patients with upper extremity disorders (excluding strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.). The authors noted that some items do not rely exclusively on upper limb function (e.g. item 9: Make a bed; item 20: manage transportation needs), and that items measure different ICF constructs (impairment, activity limitation and participationAs defined by the International Classification of Functioning, Disability and Health, participation is an individual's involvement in life situations in relation to health conditions, body functions or structures, activities, and contextual factors. Participation restrictions are problems an individual may have in the manner or extent of involvement in life situations. restriction). The authors found that patients were not able to reliably use the 5-level rating scale. Factor analysis revealed 3 underlying constructs of: (i) manual functioning (items 1-5, 7-11, 16-18, 20, 21); (ii) shoulder range of motion (items 6, 12-15, 19); and (iii) symptoms and consequences (items 22-30). Two items (Tingling, Sexual Activities) showed misfit by Rash Analysis. While results from this study identify issues to consider when using the DASH, it is important to note that patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. were excluded from the sample population.

Criterion:

Concurrent:
No studies have reported on the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the DASH in a sample of patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Predictive:
No studies have reported on the predictive validity of the DASH in a sample of patients with stroke.

Construct:

Convergent/Discriminant :
Lannin et al. (2010) conducted a comparison of the DASH with a self-report questionnaire of upper limb function and an observation upper limb movement assessment in 90 patients with stroke. The authors reported moderate correlations between manual ability and pain (statistical data not provided).

While no other studies have examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the DASH in a sample of patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., numerous studies conducted among patient groups with other upper limb conditions report adequate to excellent correlations with constructs of function and pain (see: Atroshi et al., 2000; Beaton et al., 2001; Bot et al., 2004; Kirkley et al., 1998; Schmitt & Di Fabio, 2004; SooHoo et al., 2002; Turchin et al., 1998).

Known Group:
No studies have examined known-group validityThe degree to which an assessment measures what it is supposed to measure.
of the DASH in a sample of patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., although studies have been conducted among patient groups with other upper limb conditions (see: Beaton et al., 2001).

Responsiveness

No studies have examined responsivenessThe ability of an instrument to detect clinically important change over time.
of the DASH in a sample of patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., although studies have been conducted among patient groups with other upper limb conditions (see: Beaton et al., 2001; Bot et al., 2004; MacDermid & Tottenham, 2004; Schmitt & Di Fabio, 2004).

SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
& SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
:
No studies have examined responsivenessThe ability of an instrument to detect clinically important change over time.
of the DASH in a sample of patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., although studies have been conducted among patient groups with other upper limb conditions (see: Beaton et al., 2001).

References

• Atroshi, I., Gummesson, C., Andersson, B., Dahlgren, E. & Johansson, A. (2000). The disabilities of the arm, shoulder and hand (DASH) outcome questionnaire: reliability and validity of the Swedish version evaluated in 176 patients. Acta Orthopaedica Scandinavica, 71(6), 613-8.
• Beaton, D.E., Katz, J.N., Fossel, A.H., Wright, J.G., Tarasuk, V., & Bomardier, C. (2001). Measuring the whole or the parts? Validity, reliability, and responsiveness of the Disabilities of the Arm, Shoulder and Hand outcome measure in different regions of the upper extremity. Journal of Hand Therapy, 14, 128-46.
• Beaton, D.E., Wright, J.G., Katz, J.N., and the Upper Extremity Collaborative Group. (2005). Development of the QuickDASH: comparison of three item-reduction approaches. The Journal of Bone and Joint Surgery, 87-A(5), 1038-46.
• Bot, S.D.M., Terwee, C.B., van der Windt, D.A.W.M., Bouter, L.M., Dekker, J., & de Vet, H.C.W. (2004). Clinimetric evaluation of shoulder disability questionnaires: a systematic review of the literature. Annals of the Rheumatic Diseases, 63, 335-41.
• Franchignoni, F., Biordano, A., Sartorio, F., Vercelli, S., Pascariello, B., & Ferriero, G. (2010). Suggestions for refinement of the Disabilities of the Arm, Shoulder and Hand outcome measure (DASH): a factor analysis and Rasch validation study. Archives of Physical Medicine and Rehabilitation, 91, 1370-7.
• Gummesson, C., Ward, M.M., & Atroshi, I. (2006). The shortened disabilities of the arm, shoulder and hand questionnaire (QuickDASH): validity and reliability based on responses within the full-length DASH. BMC Musculoskeletal Disorders, 7(44). doi:10.1186/1471-2474-7-44.
• Hudak, P.L., Amadio, P.C., Bombardier, C., and the Upper Extremity Collaborative Group. (1996). Development of an upper extremity outcome measure: the DASH (Disabilities of the Arm, Shoulder, and Hand). American Journal of Industrial Medicine, 29, 602-8.
• Kirkley, A., Griffin, S., McLintock, H., & Ng, L. The development and evaluation of a disease-specific quality of life measurement tool for shoulder instability: The Western Ontario Shoulder Instability Index (WOSI). The American Journal of Sports Medicine, 26(6), 764-72.
• Lannin, N. McCluskey, A. Cusick, A. Ashford, S. Ross, L. (2010) Measuring function in everyday life: enhancing the Disabilities of the Arm Shoulder Hand questionnaire for use post-stroke. World Federation of Occupational Therapy, Santiago, Chile, May.
• Lozano Calderon, S.A., Zurakowski, D., Davis, J.S., & Ring, D. (2010). Quantitative adjustment of the influence of depression on the Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire. Hand, 5, 49-55.
• MacDermid, J.C. & Tottenham, V. (2004). Responsiveness of the Disabilities of the Arm, Shoulder and Hand (DASH) and patient-rated wrist/hand evaluation (PRWHE) in evaluating change after hand therapy. Journal of Hand Therapy, 17, 18-23.
• Marx, R.G., Bombardier, C., Hogg-Johnson, S., & Wright, J.G. (1999). Clinimetric and psychometric strategies for development of a health measurement scale. Journal of Clinical Epidemiology, 52(2) 105-11.
• Ring, D., Kadzielski, J., Fabien, L., Zurakowski, D., Malhotra, L.R., & Jupiter, J.B. (2006) Self-reported upper extremity health status correlates with depression. The Journal of Bone and Joint Surgery, 88-A(9), 1983-8).
• Schmitt, J.S. & Di Fabio, R. (2004). Reliable change and minimum important difference (MID) proportions facilitated group responsiveness comparisons using individual threshold criteria. Journal of Clinical Epidemiology, 57, 1008-18.
• SooHoo, N.F., McDonald, A.P., Seiler, J.G., & McGillivrary, G.R. (2002). Evaluation of construct validity of the DASH questionnaire by correlation to the SF-36. Journal of Hand Surgery, 27A, 537-41.
• Turchin, D.C., Beaton, D.E. & Richards, R.R. (1998). Validity of observer-based aggregate scoring systems as descriptors of elbow pain, function and disability. The Journal of Bone and Joint Surgery, 80A(2), 154-62.
• Veehof, M.M., Sleegers, E.J.A., van Veldhoven, N.H.M.J., Schuurman, A.H., & van Meeteren, N.L.U. (2002). Psychometric qualities of the Dutch language version of the Disabilities of the Arm, Shoulder, and Hand questionnaire (DASH-DLV). Journal of Hand Therapy, 15, 347-54.

See the measure

How to obtain the DASH?

You can obtain a copy of the DASH through https://dash.iwh.on.ca/

Purpose

The Frenchay Arm Test (FAT) is a measure of upper extremity proximal motor control and dexterity during ADL performance in patients with impairments resulting from neurological conditions. The FAT is an upper extremity specific measure of activity limitation.

In-Depth Review

Purpose of the measure

The Frenchay Arm Test (FAT) is a measure of upper extremity proximal motor control and dexterity during ADL performance in patients with impairments of the upper extremity resulting from neurological conditions. The FAT is an upper extremity specific measure of activity limitation.

Available versions

None typically reported.

Features of the measure

Description of tasks:

Clients sit comfortably at a table with hands on their lap; each test item starts from this position. Clients are then asked to use their affected arm to:

• Stabilize a ruler, while drawing a line with a pencil held in the other hand. To pass, the ruler must be held firmly.
• Grasp a cylinder (12 mm diameter, 5 cm long), set on its side approximately 15 cm from the table edge, lift it about 30 cm and replace it without dropping.
• Pick up a glass, half full of water positioned about 15 to 30 cm from the edge of the table, drink some water and replace without spilling.
• Remove and replace a sprung clothes peg from a 10mm diameter dowel, 15 cm long set in a 10 cm base, 15 to 30 cm from table edge. Not to drop peg or knock dowel over.
• Comb hair (or imitate); must comb across top, down the back and down each side of head.

What to consider before beginning:

• Before administering the FAT, the clinician should ensure that the client is able to comprehend either written or spoken language.
• The FAT has been criticized for lacking assessment of quality of movement and performance (Kopp, 1997). In addition, clients were found to either pass or fail all or most subtests, indicating that the FAT may not be sensitive to change or subtleties in progress (Hsieh, Hsueh, Chiang & Lin, 1998), especially in clients performing in the upper range of arm function (Wade, et al., 1983).

Scoring and Score Interpretation:

Each item is scored as either pass (=1) or fail (=0). Total scores range from 0 to 5.

Time:

The FAT takes approximately 3 minutes to administer.

Training requirements:

None typically reported, however familiarity with the measure is recommended.

Equipment:

• Ruler
• Pencil
• Paper
• Cylinder (12mm diameter, 5 cm long)
• Glass (Half filled with water)
• Clothes peg
• Dowel (15mm)
• Hair comb

Alternative Forms of the FAT

None typically reported

Client suitability

Can be used with:

• Clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.

Should not be used in:

• Clients with difficulty understanding written and spoken language

Languages of the measure

• English
• French
• Dutch

Summary

What does the tool measure?	The FAT measures upper extremity proximal control and dexterity during performance of functional tasks.
What types of clients can the tool be used for?	The FAT can be used with, but is not limited to clients with stroke.
Is this a screening or assessment tool?	Assessment
Time to administer	The FAT takes approximately 3 minutes to administer.
Versions	There are no alternative versions of the FAT.
Other Languages	French and Dutch
Measurement Properties
Reliability	Intra-rater: One study examined the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the FAT in clients with stroke and found adequate to excellent intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. . Inter-rater: One study examined the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the FAT in clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and found excellent inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. .
ValidityThe degree to which an assessment measures what it is supposed to measure.	SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” / SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative). : Two studies compared the sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” of the FAT with that of the Nine-Hole Peg Test (NHPT) and found the NHPT to be more sensitive than the FAT for detecting impaired upper extremity function in clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Floor/Ceiling Effects	No studies have examined the floor/ceiling effects of the FAT in clients with stroke.
Does the tool detect change in patients?	No studies have investigated the responsivenessThe ability of an instrument to detect clinically important change over time. of the FAT in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..
Acceptability	The FAT has been criticized for lacking assessment of quality of movement and performance (Kopp, 1997). In addition, clients were found to either pass or fail all or most subtests, indicating that the FAT may not be sensitive to change (Hsieh, Hsueh, Chiang & Lin, 1998). The FAT is quick to complete and should not produce any undue fatigue for patients.
Feasibility	The FAT is short and easy to administer and score.
How to obtain the tool?	For more information on the FAT, please visit the article by Parker, Wade & Langton Hewer (1986).

Psychometric Properties

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the Frenchay Arm Test (FAT) in clients with stroke. Two studies were found and have been reviewed in this module. More studies are required before definitive conclusions can be drawn regarding the reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of "true" variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to "noise" in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
and validityThe degree to which an assessment measures what it is supposed to measure.
of the FAT.

Floor/Ceiling Effects

No studies have examined the floor/ceiling effects of the FAT in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Reliability

Internal constancy:
No studies have examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the FAT in clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Intra-rater:
Heller, Wade, Wood, Sunderland, Hewer, and Ward (1987) examined the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
of the FAT, Nine-Hole Peg Test (NHPT), Finger Tapping Rate (Lezak, 1983), and Grip Strength (Mathiowetz, Kashman, Volland, Weber, Dowe, & Rogers, 1985) in 10 patients with subacute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were re-assessed with a 2-week interval by the same rater. In this study, results describe the range of reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
of the four measures mentioned above, and values for each individual measure were not provided. Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient was adequate to excellent (ranging for all four measures from r = 0.68 to 0.99).
Note: Although is not possible to discern the exact value for the FAT reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
, all values were considered adequate to excellent and statistically significant, suggesting that the FAT may be reliable with stable strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. clients.

Inter-rater:
Heller et al. (1987) examined the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the FAT, Nine-Hole Peg Test (NHPT), Finger Tapping Rate (Lezak, 1983), and Grip Strength (Mathiowetz et al., 1985) in 10 patients with subacute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were assessed twice within a week by two raters. Spearman rho correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients were excellent (ranging for all four measures from r = 0.75 to 0.99).
Note: In this study, individual values for each measure were not provided. Although is not possible to discern the exact value for the FAT reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of "true" variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to "noise" in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
, all values were considered excellent.

Test-retest:
No studies have examined the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the FAT in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Validity

Content:

No studies have examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension.
of the FAT in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Criterion:

Concurrent:
No studies have examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the FAT in clients with stroke.

Predictive:
No studies have examined the predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the FAT in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Construct:

Convergent/Discriminant:
No studies have examined the discriminant validityMeasures that should not be related are not. Discriminant validity examines the extent to which a measure correlates with measures of attributes that are different from the attribute the measure is intended to assess.
of the FAT in clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Known Groups:
No studies have examined the known groups validityKnown groups validity is a form of construct validation in which the validity is determined by the degree to which an instrument can demonstate different scores for groups know to vary on the variables being measured.
of the FAT in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
/specificity:
Heller et al. (1987) investigated the specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
of the FAT and the Nine Hole Peg Test (NHPT) in 56 clients with chronic stroke. All of the clients that scored less than 5/5 on the FAT were correctly identified as having impaired dexterity, as identified by using the normal cut-off scores for the NHPT. However, 48 percent of patients that scored 5/5 on the FAT scored in the below normal range on the Nine Hole Peg Test. These results indicate that the NHPT is more sensitive than the FAT for detecting impaired upper extremity function in clients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Parker, Wade & Hewer (1986) compared the specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
of the FAT and the Nine-Hole Peg Test (NHPT) in 187 clients with sub-acute strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Participants that were able to successfully place nine pegs in the pegboard were further categorized according to those who completed the NHPT in less than 19 seconds (n=37) and those who required over 19 seconds (n=69). For the FAT, 114 participants score 5/5, 33 participants scored in the middle range (1/5 – 4/5) and 36 participants scored 0/5. Researchers concluded that the NHPT is more sensitive than the FAT because 13 percent of participants who scored perfectly on the FAT placed less than 9 pegs on the NHPT and all participants who scored perfectly on the NHPT (9 pegs placed in less than 19 seconds) also scored 5/5 on the FAT.

Responsiveness

No studies have examined the responsivenessThe ability of an instrument to detect clinically important change over time.
of the FAT in clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

References

• Heller, A., Wade, D.T., Wood, V.A., Sunderland, A., Langton Hewer, R., & Ward, E. (1987). Arm function after stroke: Measurement and recovery over the first three months. Journal of Neurology, Neurosurgery, and Psychiatry, 50, 714-719.
• Hsieh, C-L., Hsueh, P. Chiang, F-M., & Lin, P-H. (1998). Inter-rater reliability and validity of the Action Research Arm Test in stroke patients. Age and Ageing, 27, 107-113.
• Parker, V.M., Wade, D.T., & Langton Hewer, R. (1986). Loss of arm function after stroke: Measurement, frequency, and recovery. International Rehabilitative Medicine, 8, 69-73.
• Wade, D.T., Langton-Hewer, R., Wood, V.A., Skilbeck, C.E., & Ismail, H.M. (1983). The hemiplegic arm after stroke: Measurement and recovery. Journal of Neurology, Neurosurgery and Psychiatry, 46, 521-524.

See the measure

For more information on the FAT, please review the article by Parker, Wade & Langton Hewer (1986).

Purpose

The Jebsen Hand Function Test (JHFT) assesses fine motor skills, weighted and non-weighted hand function activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
during performance of activitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
of daily living.

In-Depth Review

Purpose of the measure

The Jebsen Hand Function Test (JHFT) is a standardized evaluative measure of functional hand motor skills (Hummel et al., 2005).

Available versions

The JHFT was developed in 1969 by Jebsen, Taylor, Treischmann, Trotter, and Howard (Cook, McCluskey, & Bowman, 2006). The JHFT is also referred to as the Jebsen-Taylor Hand Function Test or the Jebsen-Taylor Test of Hand Function.

A 3-item version (Modified Jebsen Hand Function Test, MJT) was developed by Bovend’Erdt et al. (2004) to measure gross functional dexterity in patients with moderate unilateral or bilateral upper limb impairment.

An 8-item Australian version was developed by Agnew and Maas (1982). It consists of the original 7 items with the addition of a grip strength item, measured using the Jamar dynamometer (Cook, McCluskey, & Bowman, 2006).

Features of the measure

Items:

The JHFT consists of 7 items that measure: (a) fine motor skills; (b) weighted functional tasks; and (c) non-weighted functional tasks (Jebsen et al., 1969):

• Writing a short sentence (24 letters, 3rd grade reading difficulty)
• Turning over a 3×5 inch card
• Picking up small common objects
• Simulated feeding
• Stacking checkers
• Picking up large light cans
• Picking up large heavy cans

Administration guidelines specify that testing begin with the non-dominant hand (Jebsen et al., 1969). Further details about the administration procedures of the JHFT can be found in the original article by Jebsen et al. (1969).

Items of the Modified Jebsen Hand Function Test (MJT) (Bovend’Erdt et al., 2004):

• Turning over 5 cards
• Stacking 4 cones
• Spooning 5 kidney beans into a bowl (simulated feeding)

Scoring:

Each item is scored according to time taken to complete the task. Times are rounded to the nearest second (Spinal Cord Injury Rehabilitation Evidence, 2010). The scores for all 7 items are then summed for a total score. Jebsen et al. (1969) established norms with a sample of 300 healthy subjects of different age groups (20-29 years, 30-39 years, 40-49 years, 50-59 years, 60-94 years). With the exception of writing, all items took under 10 seconds to perform. See Jebsen et al. (1969) for norms according to age, gender and hand use (dominant/non-dominant).

What to consider before beginning:

It is necessary to identify the patient’s dominant hand before beginning the JHFT. When working with patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. it is recommended to take into consideration the area(s) of cortical insult, as damage to areas of the brain responsible for speech and language function may affect performance on the writing task (Celink et al., 2007). Prior to beginning the writing task, individuals should be reminded to use reading glasses if necessary (Jebsen et al., 1969).

Time:

The JHFT requires 15 – 45 minutes to complete.

Training requirements:

No specific training is required.

Equipment:

The JHFT does not require standardized equipment but the following equipment is used (Jebsen et al., 1969):

• wooden board (41 1/2 inches long x 11 1/4 inches wide x 3/4 inch thick)
• ball point pen
• 8×11 inch sheets unruled paper
• 5×8 inch index cards
• 3×5 inch index cards
• 1 pound coffee can
• 1 inch paper clips
• teaspoon
• 5 kidney beans
• standard size wooden checkers
• 5 empty 303 cans
• 5 full (1 pound) 303 cans.

Test equipment can be collated by the clinician or purchased as pre-packaged assessment kits from suppliers including:

• Performance Health (https://www.performancehealth.com/jamar-hand-function-test)
• Mobility Smart (https://www.mobilitysmart.co.uk/jebsen-taylor-hand-function-test-kit.html)
• Amazon (amazon.com)

Client suitability

Can be used with:

• Clients with neurological or musculoskeletal conditions, e.g. strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., spinal cord injury, arthritis (Cook, McCluskey, & Bowman, 2006).
• This assessment has been administered in clients over 8 years of age (Cook, McCluskey, & Bowman, 2006).

Should not be used with:

• Individuals with speech and language disorders may have difficulty understanding instructions.
• The writing task can be excluded for individuals with speech and language difficulties due to dominant cerebral hemisphere strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (Beebe & Lang, 2009, 2007; Hummel et al., 2005).

Languages of the measure

• English
• Portuguese (Ferreiro, dos Santos, & Conforto, 2010)

Summary

What does the tool measure?	Hand function
What types of clients can the tool be used for?	The JHFT can be use with, but is not limited to clients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
Time to administer	15-45 minutes
Versions	JHFT Modified Jebsen Hand Function Test (MJT) JHFT Australian version, Portuguese version
Other Languages	English, Portuguese
Measurement Properties
ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: One study reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the JHFT (Portuguese version), and adequate to excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of individual items. Test-retest: One study reported adequate to excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of JHFT individual items. One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). of the MJT. Intra-rater: One study reported excellent intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the JHFT (Portuguese version). Inter-rater: One study reported excellent inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. of the JHFT (Portuguese version) and individual items.
ValidityThe degree to which an assessment measures what it is supposed to measure.	Content: No studies have examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension. of the JHFT. Criterion: Concurrent: Two studies reported excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between the JHFT and grip strength, pinch strength, Action Research Arm Test, Nine Hole Peg Test, and StrokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale – Hand Domain. One study reported an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between the MJT and the Nine Hole Peg Test and an adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with grip strength. Predictive: No studies have examined the predictive validityA form of criterion validity that examines a measure’s ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. of the JHFT. Construct: No studies have examined the construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed. of the JHFT. One study reported no significant difference in scores on the JHFT (Portuguese version) according to education level or hand dominance.
Floor/Ceiling Effects	No studies have examined the floor or ceiling effects of the JHFT.
SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also "Specificity." / SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).	No studies have reported on the sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.” or specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative). of the JHFT.
Does the tool detect change in patients?	One study reported moderate responsivenessThe ability of an instrument to detect clinically important change over time. of the JHFT from 1 to 3 months post-stroke, and from 3 to 6 months post-stroke.
Acceptability	The JHFT is comprised of simple, familiar, and functional tasks. Consideration must be paid to individuals with speech and language difficulties, who may have difficulty understanding instructions and performing the writing task.
Feasibility	The JHFT is easy to administer and does not require standardized equipment.
How to obtain the tool?	Information regarding test administration is provided in: Jebsen, R.H., Taylor, N., Trieschmann, R.B., Trotter, M.J., & Howard, L.A. (1969). An objective and standardized test of hand function. Archives of Physical Medicine and Rehabilitation, 50(6), 311 – 319. Assessment kits can be purchased from: Performance Health (https://www.performancehealth.com/jamar-hand-function-test) Mobility Smart (https://www.mobilitysmart.co.uk/jebsen-taylor-hand-function-test-kit.html) Amazon (www.amazon.com)

Psychometric Properties

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the Jebsen Hand Function Test (JHFT). While studies have been conducted with other patient groups, this review specifically addresses the psychometric properties relevant to patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. At the time of publication five studies were identified: three relating to the JHFT, and one each for the JHFT (Portuguese version) and the Modified Jebsen Hand Function Test (MJT).

Floor/Ceiling Effects

No studies have examined the floor or ceiling effects of the JHFT.

Reliability

Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.:
Ferreiro, dos Santos, & Conforto (2010) examined the internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the JHFT (Portuguese version) with a sample of 40 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. using Cronbach’s alpha, and reported excellent internal consistency(α=0.924). Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of individual items, reported using Pearson’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient and Cronbach’s alpha , was adequate to excellent (writing: r=0.812, α=0.844; card turning r=0.857, α=0.632; small common objects r=0.657, α=0.651; simulated feeding r=0.813, α=0.646; checkers r=0.712, α=0.633; large light objects r=0.849, α=-0.681; large heavy objects r=0.898, α=0.687).

Test-retest:
Jebsen et al. (1969) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the JHFT in a sample of 26 patients with a range of upper limb conditions including hemiparesis from cerebral vascular disease (n=5), using Pearson’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of individual tasks was adequate to excellent (writing: r=0.67, 0.84; cards: r=0.91, 0.78; small objects: r=0.93, 0.85; simulated feeding: r=0.92, 0.60; checkers: r=0.99, 0.91; large light objects: r=0.89, 0.67; large heavy objects: r=0.89, 0.92, dominant and non-dominant hands respectively).

Bovend’Eerdt et al. (2004) examined the test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the Modified Jebsen Hand Function Test (MJT) in a sample of 26 individuals with neurological disorders including strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=12), Multiple Sclerosis (n=7), head injury (n=4), and tumours (n=3). The mean time between retesting was 9.6 days. The study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MJT (r = 0.95), using Pearson’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient.

Intra-rater:
Ferreiro, dos Santos, & Conforto (2010) examined intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
of the JHFT (Portuguese version) with a sample of 40 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and reported excellent intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
(ICC=0.997), using intraclass correlation coefficient (ICC)Intraclass correlation (ICC) is used to measure inter-rater reliability for two or more raters. It may also be used to assess test-retest reliability. ICC may be conceptualized as the ratio of between-groups variance to total variance..

Inter-rater:
Ferreiro, dos Santos, & Conforto (2010) examined the inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the JHFT (Portuguese version) with a sample of 40 patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. using intraclass correlation coefficient (ICC)Intraclass correlation (ICC) is used to measure inter-rater reliability for two or more raters. It may also be used to assess test-retest reliability. ICC may be conceptualized as the ratio of between-groups variance to total variance., and reported excellent inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
(ICC=1.0). Inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
for individual items was also excellent (writing, ICC=0.999; card turning, ICC=0.977; small common objects, ICC=0.998; simulated feeding, ICC=0.991; checkers, ICC=0.995; large light objects, ICC=0.988; large heavy objects, ICC=0.991).

Validity

Content:

No studies have examined the content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension.
of the JHFT

Criterion:

Concurrent:
Beebe & Lang (2009) examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the JHFT with grip and pinch strength (measured by dynamometer), the Action Research Arm Test (ARAT) , Nine Hole Peg Test (NHPT), and the StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale – Hand domain (SIS-Hand) in a sample of 33 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
. Measures were administered at 1 month, 3 months and 6 months post-stroke. The JHFT demonstrated excellent correlations with grip strength (r=0.79-0.81), pinch strength (0.60-0.79), ARAT (r=0.87-0.95), NHPT (0.84-0.97) and SIS-Hand (0.61-0.83) at all time points.
Note: The study did not use the first task of the JHFT (writing a sentence) due to its dependence on hand dominance and education level.

Beebe & Lang (2007) examined concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the JHFT with grip and pinch strength (measured by dynamometer), Action Research Arm Test (ARAT), 9-Hole Peg Test (NHPT), and StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale – Hand Function SubscaleMany measurement instruments are multidimensional and are designed to measure more than one construct or more than one domain of a single construct. In such instances subscales can be constructed in which the various items from a scale are grouped into subscales. Although a subscale could consist of a single item, in most cases subscales consist of multiple individual items that have been combined into a composite score (National Multiple Sclerosis Society).
(SIS-Hand) in a sample of 32 participants with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., using Pearson’s product moment correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
. The JHFT demonstrated excellent correlations with ARAT (r=-0.89), grip strength (r=-0.76), pinch strength (r=-0.68), 9-HPT (r=-0.89), and SIS-Hand Function (r=-0.82).
Note: The study did not use the first task of the JHFT (writing a sentence) due to its dependence on hand dominance and education level.

Bovend’Eerdt et al. (2004) examined the concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the Modified Jebsen Hand Function Test (MJT) with the University of Maryland Arm Questionnaire for StrokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. (UMAQS), Nine Hole Peg Test (NHPT), and grip strength (measured by dynamometer) in a sample of 26 individuals with neurological disorders including strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=12), Multiple Sclerosis (n=7), head injury (n=4), and tumours (n=3). Measures were administered on two occasions (T1, T2) on average 9.6 days apart. The MJT showed excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with the NHPT (r=0.86 and 0.88 on T1 and T2 respectively) and adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with grip strength (r=0.44, significant on T2 only), using Pearson’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Correlations between the MJT and UMAQS were not significant at either time point.

Predictive:
No studies have examined the predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the JHFT.

Construct:

No studies have examined the construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the JHFT.

Known Groups:
Ferreiro et al. (2010) reported no significant difference in scores on the JHFT (Portuguese version) according to education level or hand dominance in a sample of 40 patients with strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain..

Responsiveness

Beebe & Lang (2009) measured the responsivenessThe ability of an instrument to detect clinically important change over time.
of the JHFT with a sample of 33 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using the single population effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
method. Measures were taken at 1, 3 and 6 months post-stroke, during which time participants received conventional strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. rehabilitation. The JHFT demonstrated moderate responsivenessThe ability of an instrument to detect clinically important change over time.
from 1 to 3 months post-stroke (ES=0.69) and from 3 to 6 months post-stroke (ES=0.73).

SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
& SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
:
No studies have examined the sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
and specificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
of the JHFT.

References

• Beebe, J.A. & Lang, C.E. (2007). Relating movement control at 9 upper extremity segments to loss of hand function in people with chronic hemiparesis. Neurorehabilitation and Neural Repair, 21(3), 279 – 291.
• Beebe, J.A. & Lang, C.E. (2009). Relationships and responsiveness of six upper extremity function tests during the first six months of recovery after stroke. Journal of Neurologic Physical Therapy, 33(2), 96-103.
• Bovend’Erdt, T.J.H., Dawes, H., Johansen-Berg, H., & Wade, D.T. (2004). Evaluation of the Modified Jebsen Test of Hand Function and the University of Maryland Arm Questionnaire for Stroke. Clinical Rehabilitation, 18, 195-202
• Celnik, P., Hummel, F., Harris-Love, M., Wolk, R., & Cohen, L. (2007). Somatosensory stimulation enhances the effects of training functional hand tasks in patients with chronic stroke. Archives of Physical Medicine and Rehabilitation, 88, 1369-76.
• Cook, C., McCluskey, A., & Bowman, J. (2006). Jebsen Test of Hand Function. Penrith South, NSW: University of Western Sydney. Retrieved from http://www.maa.nsw.gov.au/default.aspx?MenuID=376
• Duncan, P., Richards, L., Wallace, D., Stoker-Yates, J., Pohl, P., Luchies, C., Ogle, A., & Studenski, S. (1998). A randomized, controlled pilot study of a home-based exercise program for individuals with mild and moderate stroke. Stroke, 1998(29), 2055-2060.
• Ferreiro, K.N., dos Santos, R.L., & Conforto, A.B. (2010). Pyschometric properties of the Portuguese version of the Jebsen-Taylor test for adults with mild hemiparesis. Revista Brasileira de Fisioterapia (Brazilian Journal of Physiotherapy), 14(5), 377-81.
• Jebsen, R.H., Taylor, N., Trieschmann, R.B., Trotter, M.J., & Howard, L.A. (1969). An objective and standardized test of hand function. Archives of Physical Medicine and Rehabilitation, 50(6), 311 – 319.
• Hummel, F., Celnik, P., Giraux, P., Floel, A., Wu, W., Gerloff, C., & Cohen, L. (2005). Effects of non-invasive cortical stimulation on skilled motor function in chronic stroke. Brain, 2005(128), 490-9.
• Poole, J. (2003). Measures of Adult Hand Function: Arthritis Hand Function Test (AHFT), Grip Ability Test (GAT), Jebsen Test of Hand Function, and The Rheumatoid Hand Functional Disability Scale (The Duruöz Hand Index [DHI]). Arthritis and Rhematism (Arthritis Care and Research), 49(5S), S59-66.
• Spinal Cord Injury Rehabilitation Evidence. (2010). Jebsen Hand Function Test. Retrieved from http://www.scireproject.com/outcome-measures/jebsen-hand-function-test
• Wu, C., Seo, H., & Cohen, L. (2006). Influence of electric somatosensory stimulation on paretic-hand function in chronic stroke. Archives of Physical Medicine and Rehabilitation, 87, 351-7.

See the measure

How to obtain the JHFT?

Administration instructions are published in Jebsen, R.H., Taylor, N., Trieschmann, R.B., Trotter, M.J., & Howard, L.A. (1969). An objective and standardized test of hand function. Archives of Physical Medicine and Rehabilitation, 50(6), 311 – 319.

While the JHFT does not require standardized equipment, assessment kits can be purchased from:

• Performance Health (https://www.performancehealth.com/jamar-hand-function-test)
• Mobility Smart (https://www.mobilitysmart.co.uk/jebsen-taylor-hand-function-test-kit.html)
• Amazon (amazon.com)

Purpose

The Motor Activity Log (MAL) is a subjective measure of an individual’s real life functional upper limb performance. The MAL is administered by semi-structured interview to determine (a) how much, and (b) how well the individual uses his upper limb in his own home (Ashford et al., 2008, Li et al., 2012; Simpson & Eng, 2013).

In-Depth Review

Purpose of the measure

The Motor Activity Log (MAL) was developed by Taub et al. (1993) as a subjective outcome measure of an individual’s real life functional upper limb performance. The MAL is administered by semi-structured interview to determine (a) how much (Amount of Use – AOU), and (b) how well the individual uses his upper limb (Quality of Movement – QOM) in his own home (Ashford et al., 2008, Li et al., 2012; Simpson & Eng, 2013).

Available versions

There are four versions of the original MAL-30, according to number of items.

• MAL-14: Contains unilateral and simple items, to detect change in individuals with limited arm function.
• MAL-26: Contains the same items as the MAL-14 as well as 11 additional items and 1 optional item chosen by the patient; this version includes some bilateral tasks.
• MAL-28: Contains the same items as the MAL-14 and MAL-26, and additional items that challenge reach and strength.
• MAL-12: A short version of the MAL-28 (Ashford et al., 2008).

Other adaptations of the MAL include:

• Graded Motor Activity Log (Morera Silva et al., 2018)
• Lower-Functioning Motor Activity Log (LF-MAL)
• Lower-Extremity Motor Activity Log
• Pediatric Motor Activity Log – Revised

Features of the measure

The MAL is comprised of two scales:

• Amount of Use (AOU) scale – the amount the individual uses the paretic arm; and
• Quality of Movement (QOM) scale – the patient’s perceived quality of movement while performing the functional activity (Ashford et al., 2008).

The MAL-QOM scale captures components of amount of arm use and has been shown to be more reliable than the MAL-AOU scale, and as such can be used independently (Uswatte & Taub, 2005).

Items:

Items the original MAL-30

1. Turn on a light with a light switch
2. Open drawer
3. Remove an item from a drawer
4. Pick up phone
5. Wipe off a kitchen counter or other surface
6. Get out of a car
7. Open refrigerator
8. Open a door by turning a door knob/handle
9. Use a TV remote control
10. Wash your hands
11. Turning water on/off with knob/lever on faucet
12. Dry your hands
13. Put on your socks
14. Take off your socks
15. Put on your shoes
16. Take off your shoes
17. Get up from a chair with armrests
18. Pull chair away from table before sitting down
19. Pull chair toward table after sitting down
20. Pick up a glass, bottle, drinking cup, or can
21. Brush your teeth
22. Put on makeup base, lotion, or shaving cream on face
23. Use a key to unlock a door
24. Write on paper
25. Carry an object in your hand
26. Use a fork or a spoon for eating
27. Comb your hair
28. Pick up a cup by a handle
29. Button a shirt
30. Eat half a sandwich or finger foods

Additional Items for the MAL-45

• Removing bills from a wallet
• Taking individual coins out of a pocket or purse
• Removing keys out of a pocket or purse
• Using a zipper pull
• Pouring liquid from a bottle
• Buckling a belt
• Popping top of beverage can
• Removing top from a medicine bottle
• Keypad press
• Use of keyboard/computer
• Putting on or taking off watch band
• Putting on glasses
• Pumping a soap dispenser
• Swiping a credit card or a card for an ATM
• Adjusting a home or hotel air conditioner or heat

Items of the MAL-12:

1. Pick up phone
2. Open a door by turning a door knob
3. Eat half a sandwich or finger food
4. Turn water on/off with faucet
5. Pick up a glass
6. Pick up toothbrush and brush teeth
7. Use a key to open a door
8. Letter writing/typing
9. Use removeable computer storage
10. Pick up fork or spoon, use for eating
11. Pick up cup by handle
12. Carry an object from place to place

Items of the MAL-14:

1. Putting arm through coat sleeve
2. Steady myself while standing
3. Carry an object from place to place
4. Pick up fork or spoon, use for eating
5. Comb hair
6. Pick up cup by handle
7. Hand craft/card playing
8. Hold a book for reading
9. Use towel to dry face or other body part
10. Pick up a glass
11. Pick up toothbrush and brush teeth
12. Shaving/makeup
13. Use a key to open a door
14. Letter writing/typing

The MAL-26 includes the 14 items from the MAL-14 as well as the following items:

15. Pour coffee/tea
16. Peel fruit/potatoes
17. Dial number on the phone
18. Open/close a window
19. Open an envelope
20. Take money out of a wallet or purse
21. Undo buttons on clothing
22. Buttons on clothing
23. Undo a zip
24. Do up a zip
25. Cut fingernails (affected hand)
26. Other optional activity

Items of the MAL-28:

1. Turn on a light with a light switch
2. Open a drawer
3. Remove item of clothing from drawer
4. Pick up phone
5. Wipe kitchen counter
6. Get out of car
7. Open refrigerator
8. Open a door by turning a door knob
9. Use a TV remote control
10. Wash your hands
11. Turn water on/off with faucet
12. Dry your hands
13. Put on your socks
14. Take off your socks
15. Put on your shoes
16. Take off your shoes
17. Get up from chair with armrests
18. Pull chair away from table before sitting
19. Pull chair toward table after sitting
20. Pick up a glass
21. Pick up toothbrush and brush teeth
22. Use a key to unlock a door
23. Steady self while standing
24. Carry an object from place to place
25. Comb hair
26. Pick up cup by handle
27. Buttons on clothing (shirt, trousers)
28. Eat half a sandwich or finger food

For each item, the individual is asked whether he/she attempted the activity in the past 7 days, and the relevant score is assigned according to his/her response. The examiner can verify the response by paraphrasing it back to the individual (Uswatte & Taub, 2005). The MAL can also be used with caregivers.

Scoring:

The MAL is administered by semi-structured interview and items are scored by patients according to their performance of each task over the past 7 days; the MAL-28 can also be used to score performance over the past 3 days (Ashford et al., 2008; Uswatte & Taub, 2005).

The MAL adopts a 6-point ordinal scale, although patients can attribute a half-score, resulting in 11-point Likert scales with specified anchoring definitions at 6 points (Uswatte & Taub, 2005):

Amount of Use scale scoring:

• 0: Never – The weaker arm was not used at all for that activity.
• 1: Very rarely – Occasionally used the weaker arm, but only very rarely.
• 2: Rarely – Sometimes used the weaker arm but did the activity most of the time with the stronger arm.
• 3: Half pre-stroke – Used the weaker arm about half as much as before the strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
• 4: Three quarters pre-stroke – Used the weaker arm almost as much as before the strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..
• 5: Same – Used the weaker arm as often as before the strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Quality of Movement scale scoring:

• 0: Never – The weaker arm was not used at all for that activity.
• 1: Very rarely – The weaker arm was moved during the activity but was not very helpful.
• 2: Rarely – The weaker arm was of some use during the activity but needed some help from the stronger arm but moved very slowly or with difficulty.
• 3: Fair – The weaker arm was used for that activity, but the movements were slow or were made only with some effort.
• 4: Almost normal – The movements made by the weaker arm for the activity were almost normal but not quite as fast or accurate as normal
• 5: Normal – The ability to use the weaker arm for that activity was as good as before the strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain..

Scale total scores (summary scores) are the mean of the item scores.

What to consider before beginning:

The MAL is subject to experimenter bias and also the patient’s ability to accurately recall upper limb use (Page & Levine, 2003; Uswatte & Taub, 2005).

Ashford et al. (2008) noted an inadequate relationship between overall/item scores and the qualitative meaning, and an unclear Minimal clinically important difference.

Taub & Uswatte (2000) discuss the use of the MAL as an outcome measure in Constraint-Induced Movement Therapy (CIMT) research and recommend an upper cut-off score of 2.5 on the MAL-AOU, as the effect of strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. can impose an upper physiological limit on the amount of improvement that can be produced. The authors also note that individuals who score > 2.5 do not demonstrate learned non-use, which is the aim of CIMT.

Time:

All versions of the MAL are administered through structured interview with the patient and/or carer and require more than 10 minutes to administer. (Ashford et al., 2008).

Training requirements:

The MAL can be administered by health professionals who have reviewed the manual and literature.

Equipment:

Survey instrument and pencil.

Client suitability

Can be used with:

• The MAL is suitable for use with adults and elderly adults following strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers. It is suitable for use in the subacute and chronic stages of strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. recovery.

Should not be used in:

Not specified.

• The MAL is often used to measure outcomes following constraint induced movement therapy (Li et al., 2012; Page, 2003). The MAL is commonly used in research in conjunction wi with the Wolf Motor Function Test, Fugl-Meyer Assessment or the Action Research Arm Test (Santisteban et al., 2016; Simpson & Eng, 2013).

In what languages is the measure available?

• Brazilian-Portuguese (Saliba et al., 2011)
• English
• German (Khan et al., 2013)
• Portuguese (Pereira et al., 2011)
• Turkish translation and cultural adaptation (Cakar et al., 2010).

Summary

What does the tool measure?	Real life upper limb performance.
What types of clients can the tool be used for?	Individuals following strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers.
Is this a screeningTesting for disease in people without symptoms. or assessment tool?	Assessment
What domain of the ICF does this measure?	Activity/participation
Time to administer	20 minutes
Versions	MAL-30 MAL-28 MAL-26 MAL-14 MAL-12 Graded Motor Activity Log Lower-Functioning Motor Activity Log (LF-MAL) Lower-Extremity Motor Activity Log Pediatric Motor Activity Log – Revised
Other Languages	Brazilian-Portuguese, English, German, Portuguese, Turkish.
Measurement Properties
Reliability	Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.: – MAL-14: Two studies reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.. – MAL: One study reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.; one study reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. among patients with mild-moderate hemiparesis and adequate to excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. among patients with severe hemiparesis. – MAL-28 (Turkish): One study reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.. – MAL-30 (German): One study reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.. – Grade 4/5 MAL: One study reported excellent internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.. Test-retest: – MAL-14: One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). ; one study reported adequate to excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). . – MAL: One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). ; one study reported adequate to excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). . – MAL-28 (Turkish): One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). . – MAL-28 (Brazilian): One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). . – MAL-45: One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). . – Grade 4/5 MAL: One study reported excellent test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society). . Intra-rater: No studies have reported on the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings. of the MAL. Inter-rater: MAL-14: One study reported adequate inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept. .
ValidityThe degree to which an assessment measures what it is supposed to measure.	Content: No studies have reported on content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension. of the MAL. Criterion: Concurrent: – MAL-14: One study reported excellent correlations with accelerometry. – MAL: Three studies reported an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. with SIS – Hand function domain; adequate correlations with the BBT, ARAT, FAI; poor to adequate correlations with SIS, SS-QOL, NEADL; and poor correlations with the Nine Hole Peg Test. – MAL-30 (German): One study reported excellent negative correlations with WMFT-PT; excellent correlations with WMFT-FA and Grip strength scores, CMSA – Arm and Hand scores, isometric strength. – MAL-45: 1 study reported excellent correlations with the Abilhand. Predictive: No studies have reported on predictive validityA form of criterion validity that examines a measure’s ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks. of the MAL. Construct: – MAL-14: One study reported excellent correlations between QOM and AOU patient/carer change scores; one study reported an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between AOU and QOM scales. – MAL: One study reported an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between AOU and QOM scales; one study reported an adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between AOU and QOM scales; one study conducted item analysis and removed two items due to low item-total correlations and reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest . coefficients; one study conducted item fit analysis and principal component analysis. – MAL (Brazilian): One study reported an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between AOU and QOM scales. – MAL-30 (German): One study reported excellent correlations between AOU and QOM scales. – MAL-28 (Turkish): One study reported an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between AOU and QOM scales. – LF-MAL: One study reported an adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation. between the AOU and QOM scales. Convergent/Discriminant: – MAL-14: Three studies reported excellent correlations with ARAT, accelerometry, Simple Test for Evaluating Hand Function (STEF). – MAL: Seven studies reported excellent correlations with Actual Amount of Use Test, WMFT; adequate to excellent correlations with accelerometry ratios, SIS 2.0 – Hand function scale, FMA-UE; adequate correlations with ARAT, Motor Assessment Scale – Upper Extremity, 16 Hole Peg Test, grip strength; SF-36 – Physical domain; poor to adequate correlations with accelerometry ratios of the less affected arm; poor correlations with the SIS 2.0 – Mobility scale. – MAL-28 (Turkish): One study reported excellent correlations with WMFT-FA; adequate negative correlations with the WMFT-PT. – MAL (Brazilian): One study reported adequate correlations with grip strength of the more affected arm. Known Group: MAL: One study reported correlations with accelerometry was stronger among patients with paresis of the dominant arm vs. the non-dominant arm.
Floor/Ceiling Effects	– Floor effects are evident when detecting change in lower level and passive functional tasks. – One study found modest floor effects when the MAL-28 was administered to patients with upper extremity motor recovery at Brunnstrom stage III and higher; and modest floor effects when the LF-MAL was administered to patients with upper extremity motor recovery at Brunnstrom stage III and lower.
Does the tool detect change in patients?	The MAL can be used to detect change
Acceptability	The MAL reflects real life functional performance. It is simple and non-invasive to administer.
Feasibility	The MAL is a free tool that requires no additional equipment. It can be administered in the clinical setting or the patient’s home. No additional training is required.
How to obtain the tool?	Click here to see the Motor Activity Log manual.

Psychometric Properties

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the MAL. Twenty-six studies were identified, most of which included patients in the chronic phase of strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. recovery. This review includes different versions of the MAL – the original MAL-30, MAL-28, MAL-14, MAL-45, LF-MAL, Grade 4/5 MAL and Turkish, Brazilian and German versions.

Floor/Ceiling Effects

Chuang et al. (2017) examined floor/ceiling effects of the 30-item MAL in a sample of 403 patients with chronic stroke. The MAL was administered to patients with motor recovery of the proximal and distal upper limb at Brunnstrom stage III and higher. Results showed modest floor effects within this cohort, whereby 17.3% of participants received minimum scores on the MAL.

Chuang et al. (2017) examined floor/ceiling effects of the LF-MAL in a sample of 134 patients with chronic stroke. The LF-MAL was administered to patients with motor recovery of the proximal and distal upper limb at Brunnstrom stage III and lower. Results showed modest floor effects within this cohort, whereby 16.4% of participants received minimum scores on the LF-MAL.

Reliability

Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency.:
van der Lee et al. (2004) examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-14 in a sample of 56 patients with chronic stroke, using Cronbach’s alpha. Correlation among items was excellent for the MAL-AOU (a = 0.87) and the MAL-QOM (a = 0.90). Limits of agreement ranged from -0.70 to 0.85 for the MAL-AOU and from -0.61 to 0.71 for the MAL-QOM, indicating reproducibility sufficient to detect an individual change of approximately 12-15% of the range of the scale.

Uswatte et al. (2005b) examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-14 in a sample of 41 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers, using Cronbach’s alpha. CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
among items was excellent for patients’ MAL-QOM (a = 0.87) and caregivers’ MAL-AOU and MAL-QOM (a > 0.83). The authors also examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-14 (QOM scale only) in a sample of 27 patients with chronic stroke. Correlation among items was excellent for the MAL-QOM (a = 0.81).

Uswatte et al. (2006b) examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-28 in a sample of 222 patients with subacute/chronic stroke and their caregivers, using Cronbach’s alpha. Responses from both patient and caregiver groups showed excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
among items for the MAL-AOU (patients a = 0.94; caregivers a = 0.95) and the MAL-QOM (patients a = 0.94; caregivers a = 0.95).

Huseyinsinoglu et al. (2011) examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-28 (Turkish version) in a sample of 30 patients with stroke, using Cronbach’s alpha. Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. was excellent for the MAL-AOU (a = 0.96) and MAL-QOM (a = 0.96).

Khan et al. (2013) examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-30 (German version) in a sample of 42 patients with acute to chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain., using Cronbach’s alpha. Measures were taken at baseline, discharge from rehabilitation and at 6-month follow-up. Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. for the MAL-AOU and MAL-QOM were excellent at all timepoints (a = 0.98-0.995). The authors also calculated internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. based on an elimination procedure of items that scored “N/A” down to 26 items and reported that internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. remained high at all timepoints (a = 0.94-0.98).

Taub et al. (2013) reported on internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the Grade 4/5 MAL, referencing unpublished data from Morris (2009) that used a sample of 30 individuals with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Cronbach’s alpha. Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. for the Grade 4/5 MAL was excellent (a = 0.95).

Chuang et al. (2017) examined the 6-point rating system of the MAL and found rater difficulty discriminating among the 6 levels of functional ability. Results showed that 15 items of the MAL-AOU and MAL-QOM displayed disordering of step difficulty. Accordingly, the 6 levels were collapsed into 4 levels to restore reversed threshold (0 = 0; 1-2 = 1; 3-4 = 2; 5 = 3); using the 4-point system 9 items still showed disordered ordering, so the levels were further collapsed into 2 categories (0 = 0; 1 to 3 = 1), at which point all items exhibited ordering. The authors examined unidimensionality of the 30-item MAL in a sample of 403 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using the revised scoring system. Item fit analysis of the MAL revealed that 7 items* of the MAL-AOU and MAL-QOM were a poor fit and were removed. Principal component analysis (PCA) of the remaining 23 items showed that Rasch measures accounted for 76% of the variance for both the MAL-AOU and MAL-QOM, with an eigenvalue of the first residual factor of 2.7. This indicates that the 23 items constitute unidimensional constructs. The authors examined reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of "true" variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to "noise" in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
of the revised MAL (23 items, 4-point rating system), using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
. With Pearson separation values of 2.4 and 2.6 for the MAL-AOU and MAL-QOM respectively, the revised version was sensitive to distinguish among 3 strata of upper limb performance. Pearson reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
coefficients were 0.85 and 0.87 (respectively), suggesting good reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
. Results showed no Differential Item Functioning (DIF) items across age, gender or hand dominance. Item difficulty hierarchy was consistent with clinical expectation, however items were more difficult than individuals’ ability, suggesting unsuitable targeting for the participants of this sample.

* Misfit items: (6) Get out of car; (12) Dry your hands; (18) Pull a chair away from the table before sitting down; (19) Pull chair toward table after sitting down; (21) Brush your teeth; (24) Write on paper; (29) Button a shirt.

Chuang et al. (2017) examined the 6-point rating system of the LF-MAL and found disordered thresholds; accordingly, the 6 levels were collapsed into 3 levels to restore reversed threshold (0 = 0; 1-3 = 1; 4-5 = 2); this 3-point rating system achieved step ordering. The authors examined unidimensionality of the LF-MAL in a sample of 134 patients with chronic stroke, using the revised 3-point scoring system. Item fit analysis of the LF-MAL-AOU revealed that 6 items were out of the acceptable range; PCA of the remaining 24 items showed that the Rasch dimension explained 70.5% of the variance, with an eigenvalue of 2.6 of the first residual factor. Item fit analysis of the LF-MAL-QOM revealed that 7 items were out of the acceptable range; PCA of the remaining 23 items showed that the Rasch dimension explained 71.0% of the variance, with an eigenvalue of the first residual factor of 2.5. The authors examined reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
of the revised LF-MAL (25 items, 3-point rating system), using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
. With Pearson separation values of 1.9 for both the LF-MAL-AOU and LF-MAL-QOM, the revised version was sensitive to distinguish 2 strata of upper limb performance. Pearson reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
coefficients were 0.79 for both the LF-MAL-AOU and LF-MAL-QOM, indicating acceptable reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
. Results showed no DIF items across age, gender or hand dominance. Item difficulty hierarchy was consistent with clinical expectation, however items were more difficult than individuals’ ability, suggesting unsuitable targeting for the participants of this sample.

* Misfit items: (5) Wipe off a kitchen counter or another surface; (6) Get out of a car; (7) Open a refrigerator; (19) Apply soap to your body while bathing (LF-MAL-QOM only); (21) Brush your teeth; (23) Steady yourself while standing; (24) Carry an object in your hand.

Moreira Silva et al. (2018) examined internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-30 in a sample of 66 individuals with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Cronbach’s alpha. Participants were classified according to upper extremity motor function using the Fugl-Meyer Assessment – Upper Extremity (FMA-UE): mild to moderate hemiparesis (FMA-UE ≥ 32, n = 49) or severe hemiparesis (FMA-UE ≤31, n = 17). Internal consistencyA method of measuring reliability . Internal consistency reflects the extent to which items of a test measure various aspects of the same characteristic and nothing else. Internal consistency coefficients can take on values from 0 to 1. Higher values represent higher levels of internal consistency. of the MAL-AOU and MAL-QOM was excellent among participants with mild-moderate hemiparesis (a = 0.95), and adequate to excellent among participants with severe hemiparesis (MAL-AOU: a = 0.79; MAL-QOM: a = 0.89). Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
was used to further evaluate reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
of the MAL-30. Item calibration of the MAL-AOU and MAL-QOM revealed one misfit (#19: Pull a chair toward table after sitting down). Item separation index of the MAL-AOU and MAL-QOM was 2.92 and 2.59 (respectively) suggesting 5 levels of difficulty for the MAL-AOU and 4 levels of difficulty for the MAL-QOM. Pearson separation index of the MAL-AOU and MAL-QOM was 2.62 and 2.58 (respectively), suggesting 4 ability levels for both the MAL-AOU and the MAL-QOM.

Test-retest:
Miltner et al. (1999) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL in a sample of 15 patients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain.. Measures were taken within a 2-week interval before participants began constraint-induced movement therapy. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
was excellent (r = 0.98).

Johnson et al. (2003) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL-45 in a sample of 12 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Pearson’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Measures were taken within a 3-week interval. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
was excellent for the MAL-AOU (r=0.96) and MAL-QOM (r = 0.99).

van der Lee et al. (2004) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL-14 in a sample of 56 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using the Bland and Altman method. Measures were taken within a 2-week interval before participants commenced an intervention program. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
was excellent for the for MAL-AOU (r = 0.70 to 0.85) and the MAL-QOM (r = 0.61 to 0.71).

Uswatte et al. (2005b) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL-14 in a sample of 41 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers, using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
was excellent for patient MAL-QOM scores (r = 0.91), and adequate for patient MAL-AOU scores (r = 0.44), and caregiver MAL-AOU and MAL-QOM scores (r = 0.61, r = 0.50 respectively).

Uswatte et al. (2006b) examined 2-week test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL-30 in a sample of 116 patients with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers, using Intra Class Coefficients (ICC). Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
for the MAL-AOU and MAL-QOM was excellent among patients (ICC = 0.79, ICC = 0.82, respectively), and adequate among caregivers (ICC = 0.66, ICC = 0.72, respectively). There was a trend toward an increase from test 1 to test 2 among both patients and caregivers (patient MAL-AOU: 0.3 ± 0.6, p = 0.04; patient MAL-QOM: 0.3 ± 0.5, p = 0.02; caregiver MAL-AOU: 0.4 ± 0.7, p = 0.05; caregiver MAL-QOM: 0.4 ± 0.7, p = 0.02), although increases were less than the minimal clinically important difference (< 0.5 points).

Huseyinsinoglu et al. (2011) examined 3-day test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL-28 (Turkish version) in a sample of 30 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using intraclass coefficients (ICC) and Spearman correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
was excellent for the MAL-AOU (ICC = 0.97, r = 0.94) and the MAL-QOM (ICC = 0.96, r = 0.93).

Saliba et al. (2011) examined test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the MAL (Brazilian version), using intra-class correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients (ICC). Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
for the MAL-AOU and MAL-QOM was excellent (ICC = 0.98).

Taub et al. (2013) reported on test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
of the Grade 4/5 MAL, referencing unpublished data from Morris (2009) that used a sample of 10 individuals with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Test-retest reliabilityA way of estimating the reliability of a scale in which individuals are administered the same scale on two different occasions and then the two scores are assessed for consistency. This method of evaluating reliability is appropriate only if the phenomenon that the scale measures is known to be stable over the interval between assessments. If the phenomenon being measured fluctuates substantially over time, then the test-retest paradigm may significantly underestimate reliability. In using test-retest reliability, the investigator needs to take into account the possibility of practice effects, which can artificially inflate the estimate of reliability (National Multiple Sclerosis Society).
for the Grade 4/5 MAL was excellent (r = 0.95).

Intra-rater:
No studies have reported on the intra-rater reliabilityThis is a type of reliability assessment in which the same assessment is completed by the same rater on two or more occasions. These different ratings are then compared, generally by means of correlation. Since the same individual is completing both assessments, the rater’s subsequent ratings are contaminated by knowledge of earlier ratings.
of the MAL.

Inter-rater:
Uswatte et al. (2005b) examined inter-rater reliabilityA method of measuring reliability . Inter-rater reliability determines the extent to which two or more raters obtain the same result when using the same instrument to measure a concept.
of the MAL-14 in a sample of 41 patients with chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers using Intra Class Coefficients (ICC). Participants received Constraint-Induced Movement Therapy (CIMT) or time-matched general fitness rehabilitation for two weeks. ReliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of "true" variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to "noise" in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
between patient and carer pre-treatment scores was adequate (ICC = 0.52, p < 0.01); reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of "true" variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to "noise" in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
between patient and carer change scores following treatment was adequate (ICC = 0.7, p < 0.0001).

Validity

Content:

No studies have reported on content validityRefers to the extent to which a measure represents all aspects of a given social concept. Example: A depression scale may lack content validity if it only assesses the affective dimension of depression but fails to take into account the behavioral dimension.
of the MAL.

Criterion:

Concurrent:
Johnson et al. (2003) examined concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the MAL-45 in a sample of 12 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with the Abilhand, using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Correlations with the Abilhand were excellent for the MAL-AOU (r = 0.71, p < 0.05) and MAL-QOM (r = 0.88, p < 0.05).

Uswatte et al. (2005b) examined concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the MAL-14 (QOM scale only) in a sample of 27 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with accelerometry of the affected arm, using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Correlations between the MAL-QOM and accelerometer recordings at pre-treatment (r = 0.70, p < 0.05) were excellent. Correlations between MAL-QOM change scores from pre-treatment to post-treatment and corresponding change scores on accelerometer readings were also excellent (r = 0.91, p < 0.01).

Lin et al. (2010a) examined concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the MAL-30 by comparison with the Nine Hole Peg Test (9HPT), the Box and Block Test (BBT) and the Action Research Arm Test (ARAT), using Spearman rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n=59) were randomized to receive distributed constraint-induced movement therapy, bilateral arm training or neurodevelopmental therapy, and measures were taken at baseline and post-treatment (3 weeks). Correlations at baseline and post-treatment were significant and adequate with the BBT (MAL-AOU: r = 0.37, r = 0.49; MAL-QOM: r = 0.52, r = 0.52) and the ARAT (MAL-AOU: r = 0.31, r = 0.32; MAL-QOM: r = 0.39, r = 0.35). Correlations with the 9HPT were significant for the MAL-QOM only (r = -0.26, r = -0.33).

Lin et al. (2010b) examined concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the MAL-30 by comparison with the StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale 3.0 (SIS) and the Stroke-Specific Quality of Life Scale (SS-QOL), using Spearman rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n = 74) were randomized to receive distributed constraint-induced movement therapy, bilateral arm training or neurodevelopmental therapy, and measures were taken at baseline and post-treatment (3 weeks). There were significant poor to adequate correlations between the MAL-AOU and most SIS domains at baseline (r = 0.24-0.58) and post-treatment (r = 0.24-0.59). There were significant excellent correlations between the MAL-QOM and the SIS – Hand function domain at baseline (r = 0.65) and post-treatment (r = 0.68), and significant poor to adequate correlations between the MAL-QOM and most other SIS domains at baseline (r = 0.26-0.52) and post-treatment (r = 0.28-0.51). There were significant correlations between the MAL-AOU and some SS-QOL domains at baseline (r = 0.25-0.37) and post-treatment (r = 0.24-0.35), and between the MAL-QOM and some SS-QOL domains at baseline (r = 0.28-0.38) and post-treatment (r = 0.26-0.39).

Wu et al. (2011) examined concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the MAL-30 in a sample of 77 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with a modified version of the Nottingham Extended ADL Scale (NEADL) and the Frenchay ActivitiesAs defined by the International Classification of Functioning, Disability and Health, activity is the performance of a task or action by an individual. Activity limitations are difficulties in performance of activities. These are also referred to as function.
Index (FAI), using Spearman rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Measures were taken at pre-treatment and 3 weeks later at post-treatment. Correlations with the NEADL were poor to adequate (MAL-AOU: r = 0.3; MAL-QOM: r = 0.2-0.3). Correlations with the FAI were adequate (MAL-AOU: r = 0.3-0.4); MAL-QOM: r = 0.3).

Khan et al. (2013) examined cross-sectional concurrent validityTo validate a new measure, the results of the measure are compared to the results of the gold standard obtained at approximately the same point in time (concurrently), so they both reflect the same construct. This approach is useful in situations when a new or untested tool is potentially more efficient, easier to administer, more practical, or safer than another more established method and is being proposed as an alternative instrument. See also “gold standard.”
of the MAL-30 (German version) by comparison with the Wolf Motor Function Test (WMFT) – Time and Functional ability subtests, the Chedoke McMaster StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Assessment (CMSA) – Arm and Hand subtests, the grip strength scale, and isometric strength measured by handheld dynamometer (mean of shoulder and elbow flexion and extension), using Spearman’s rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Patients with acute to chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n = 42) received inpatient rehabilitation and measures were taken at baseline; discharge from hospital and at 6-month follow-up. Significant negative correlations were seen with the WMFT – Time scores (MAL-AOU r = -0.747 – -0.878; MAL-QOM r = -0.770 – -0.901). Correlations were excellent at all time points with the WMFT – Functional ability (MAL-AOU r = 0.769 – 0.808, MAL-QOM r = 0.789 – 0.837), the CSMA – Arm (MAL-AOU r = 0.680 – 0.765; MAL-QOM r = 0.691 – 0.798) and CSMA – Hand (MAL-AOU r = 0.692 – 0.801; MAL-QOM r = 0.717 – 0.803), grip strength (MAL-AOU r = 0.698 – 0.716; MAL-QOM r = 0.659-.0733) and isometric strength (MAL-AOU r = 0.643-0.719; MAL-QOM r = 0.714-0.726).

Predictive:
No studies have examined predictive validityA form of criterion validity that examines a measure's ability to predict some subsequent event. Example: can the Berg Balance Scale predict falls over the following 6 weeks? The criterion standard in this example would be whether the patient fell over the next 6 weeks.
of the MAL.

Construct:

Uswatte et al. (2006b) conducted item analysis of the original MAL-30 using item-total correlations, reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
and proportion of missing data (with an a priori cut-off of 20%) in a sample of 222 patients with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers. Of the 30 items, 25 items were completed by > 80% of caregivers and 28 items were completed by > 80% of patients; analysis of these 28 items indicated item-total correlations > 0.5 for 92% of items, and reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
coefficients > 0.5 for 89% of items. The remaining 2 items (write on paper: 48% missing data; put makeup/shaving cream on face: 20% missing data) showed lower item-total correlations and reliabilityReliability can be defined in a variety of ways. It is generally understood to be the extent to which a measure is stable or consistent and produces similar results when administered repeatedly. A more technical definition of reliability is that it is the proportion of “true” variation in scores derived from a particular measure. The total variation in any given score may be thought of as consisting of true variation (the variation of interest) and error variation (which includes random error as well as systematic error). True variation is that variation which actually reflects differences in the construct under study, e.g., the actual severity of neurological impairment. Random error refers to “noise” in the scores due to chance factors, e.g., a loud noise distracts a patient thus affecting his performance, which, in turn, affects the score. Systematic error refers to bias that influences scores in a specific direction in a fairly consistent way, e.g., one neurologist in a group tends to rate all patients as being more disabled than do other neurologists in the group. There are many variations on the measurement of reliability including alternate-forms, internal consistency , inter-rater agreement , intra-rater agreement , and test-retest .
coefficients and were dropped accordingly.

van der Lee et al. (2004) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-14 in a sample of 56 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There was an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-AOU and MAL-QOM (r = 0.95, p < 0.001).

Uswatte et al. (2005b) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-14 (QOM scale only) in a sample of 27 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with patient/caregiver MAL-AOU scores, using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Correlations were excellent between MAL-QOM change scores from pre-treatment to post-treatment and corresponding change scores in patient MAL-AOU (r = 0.80, p < 0.01), carer MAL-AOU (r = 0.73, p < 0.01) and carer MAL-QOM (r = 0.70, p < 0.01).

Uswatte et al. (2006a) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-30 in a sample of 169 individuals with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There was an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-AOU and MAL-QOM (r = 0.92, p < 0.001).

Huseyinsinoglu et al. (2011) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-28 (Turkish version) in a sample of 30 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. The correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-AOU and the MAL-QOM was excellent (r = 0.95).

Saliba et al. (2011) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL (Brazilian version) in a sample of 77 individuals with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
. There was an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-AOU and the MAL-QOM (r = 0.97, p < 0.0001).

Khan et al. (2013) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-30 (German version), using Spearman’s rank correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Patients with acute to chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n = 42) received inpatient rehabilitation and measures were taken at baseline, discharge from hospital and at 6-month follow-up. There was an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-AOU and MAL-QOM at all timepoints (r = 0.994, 0.982, 0.980).

Chuang et al. (2017) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-30 in a sample of 403 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. with motor recovery of the proximal and distal upper limb at Brunnstrom stage III and higher, using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
. CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-AOU and MAL-QOM was adequate (r = 0.603), indicating that the subscales are not highly correlated and can be perceived as different concepts.

Chuang et al. (2017) examined construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the LF-MAL in a sample of 134 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. with motor recovery of the proximal and distal upper limb at Brunnstrom stage III and lower, using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
. CorrelationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the LF-MAL-AOU and LF-MAL-QOM was adequate (r = 0.607), indicating that the subscales are not highly correlated and can be perceived as different concepts.

Convergent/Discriminant:
van der Lee et al. (2004) examined cross-sectional convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-14 by comparison with the Action Research Arm Test (ARAT) in a sample of 56 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There were excellent correlations between the MAL-AOU and the ARAT (r = 0.63, p < 0.001) and between the MAL-QOM and the ARAT (r = 0.63, p < 0.001).

Uswatte et al. (2005a) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-14 in a sample of 20 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with accelerometry of the affected arm, using Spearman rank correlations. There was an excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-14 and accelerometry (r = 0.74, p < 0.001).

Uswatte et al. (2006a) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-30 (QOM scale only) in a sample of 169 patients with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with accelerometry of the affected arm and the Actual Amount of Use Test (AAUT), using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases - for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases - for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Correlations between the MAL-QOM and accelerometry ratios (ratio summary variable, impaired arm summary variable) were adequate (r = 0.52, r = 0.41 respectively, p < 0.001). The correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the MAL-QOM and AAUT was excellent (r = 0.94, p < 0.001).

Uswatte et al. (2006b) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-30 in a sample of 222 patients with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers by comparison with accelerometry of the affected arm, and the SIS 2.0 – Hand function scale, using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Comparison of the MAL with accelerometry ratios showed adequate to excellent correlations for patient scores (MAL-AOU: r = 0.47; MAL-QOM: r = 0.52, p < 0.01), and adequate correlations for caregiver scores (MAL-AOU: r = 0.57; MAL-QOM, r = 0.61, p < 0.01). Comparison of the MAL and SIS – Hand function scores showed excellent correlations for patient scores (MAL-AOU: r = 0.68; MAL-QOM: r = 0.72, p < 0.01), and adequate correlations for caregiver scores (MAL-AOU: r = 0.35, MAL-QOM: r = 0.40, p < 0.01).

Uswatte et al. (2006b) examined divergent validityThe degree to which an assessment measures what it is supposed to measure.
of the MAL-30 in a sample of 222 patients with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers by comparison with accelerometry of the less affected arm, and the SIS 2.0 – Mobility scale, using Pearson correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficients. Comparison of the MAL with accelerometry ratios of the less affected arm showed poor correlations for patient scores (MAL-AOU: r = 0.14; MAL-QOM: r = 0.14, p > 0.05), and poor to adequate correlations for caregiver scores (MAL-AOU: r = 0.25; MAL-QOM, r = 0.23, p < 0.001). Comparison of the MAL and SIS – Mobility scores showed poor correlations for patient scores (MAL-AOU: r = 0.14; MAL-QOM: r = 0.14, p > 0.05), and poor correlations for caregiver scores (MAL-AOU: r = 0.10, MAL-QOM: r = 0.07, p > 0.05).

Hammer and Lindmark (2010) examined cross-sectional convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-30 by comparison with the FMA-UE, ARAT, Motor Assessment Scale – Upper Extremity score (MAS-UE), 16-hole peg test (16HPT) and the Grippit ratio of isometric grip strength, using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Patients with subacute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. (n = 30) were randomized to receive forced use therapy or standard upper limb rehabilitation, and measures were taken at baseline, post-treatment (2 weeks) and follow-up (3 months). Correlations were significant and adequate with all measures: FMA-UE (r = 0.43-0.52); ARAT (r = 0.31-0.51); MAS-UE (r = 0.41-0.54); 16HPT (r = -0.41 – -0.67); Grippit (r = 0.41-0.53).

Huseyinsinoglu et al. (2011) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-28 (Turkish version) by comparison with the WMFT – Performance Time (WMFT-PT) and – Functional Ability (WMFT-FA) scores in a sample of 30 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. There were excellent correlations with the WMFT-FA (MAL-AOU, r=0.63; MAL-QOM: r = 0.63), and adequate negative correlations with the WMFT-PT (MAL-AOU: r = -0.56; MAL-QOM: r = -0.55).

Saliba et al. (2011) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL (Brazilian version) by comparison with grip strength of the more severely affected upper limb in a sample of 77 individuals with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using Rasch analysisRasch analysis is a statistical measurement method that allows the measurement of an attribute – such as upper limb function – independently of particular tests or indices.   It creates a linear representation using many individual items, ranked by item difficulty (e.g. picking up a very small item, versus a task requiring a very gross grasp) and person ability.   A well performing Rasch model will have items hierarchically placed from simple to more difficult, and individuals with high abilities should be able to perform all the items below a level of difficulty.  The Rasch model is statistically strong because it enables ordinal measures to be converted into meaningful interval measures. It also allows information from various tests or tools with different scoring systems to be applied using the Rasch model.
. There were adequate correlations between grip strength and the MAL-AOU (r = 0.51, p < 0.0001) and the MAL-QOM (r =0 .57, p < 0.0001).

Sterr et al. (2014) examined divergent validityThe degree to which an assessment measures what it is supposed to measure.
of the MAL in a sample of 65 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with the Short Form 36 (SF-36), StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale (SIS), Hospital Anxiety and DepressionIllness involving the body, mood, and thoughts, that affects the way a person eats and sleeps, the way one feels about oneself, and the way one thinks about things. A depressive disorder is not the same as a passing blue mood or a sign of personal weakness or a condition that can be wished away. People with a depressive disease cannot merely “pull themselves together” and get better. Without treatment, symptoms can last for weeks, months, or years. Appropriate treatment, however, can help most people with depression.
Scale (HADS) and Visual Analog Mood Score (VAMS), using regression analysis. Participants received four different Constraint-Induced Movement Therapy (CIMT) treatment protocols that differed in intensity and use of a constraint. Following treatment there was a significant positive association between the MAL-AOU and the SF-36 Physical domain (r = 0.38m p = 0.025) and a trend towards a moderate association with the SIS Total score (r = 0.43, p = 0.061).

Shindo et al. (2015) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-14 in a sample of 34 patients with acute/subacute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with the Simple Test for Evaluating Hand Function (STEF), using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There was a significant and excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
between the assessments (MAL-AOU: r = 0.805; MAL-QOM: r = 0.768).

Simpson, Conroy & Beaver (2015) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-28 in a sample of 9 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., by comparison with the FMA, Wolf Motor Function Test and StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Impact Scale, using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There were excellent correlations between baseline MAL-AOU and FMA (ρ = 0.6889, p < 0.0132) and MAL-QOM and FMA (ρ = 0.7276, p < 0.0073).

Moreira Silva et al. (2018) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL-30 in a sample of 66 individuals with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with the FMA-UE, using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There was a significant and excellent correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with the FMA-UE (MAL-AOU: r = 0.87; MAL-QOM: r = 0.87).

Chen et al. (2018) examined convergent validityA type of validity that is determined by hypothesizing and examining the overlap between two or more tests that presumably measure the same construct. In other words, convergent validity is used to evaluate the degree to which two or more measures that theoretically should be related to each other are, in fact, observed to be related to each other.
of the MAL in a sample of 82 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. by comparison with accelerometry of the affected arm, using Pearson’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. There was an adequate correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
with accelerometry (MAL-AOU: r = 0.47; MAL-QOM: r = 0.57).

Known Group:
Uswatte et al. (2006b) examined known-group validityThe degree to which an assessment measures what it is supposed to measure.
of the MAL in a sample of 222 patients with subacute/chronic strokeAlso called a "brain attack" and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a "schemic stroke", or the formation of a blood clot in a vessel supplying blood to the brain. and their caregivers. Correlations between the MAL and accelerometry ratio was stronger among patients with paresis of their dominant arm (MAL-AOU: r = 0.56; MAL-QOM: r = 0.59) than among patients with paresis of the non-dominant arm (MAL-AOU: r = 0.28; MAL-QOM: r = 0.34).

Responsiveness

Taub et al. (1993) reported on Effect sizes (ES) of the MAL in a sample of 9 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants received two weeks of upper extremity restraint and measures were taken at baseline, post-treatment and follow-up (1 month, 2 years). Effect sizes were large from baseline to 1-month follow-up (2.80) and from baseline to 2-year follow-up (2.95).

Kunkel et al. (1999) reported on ES of the MAL in a sample of 5 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants received two weeks of Constraint-Induced Movement Therapy (CIMT) and measures were taken at baseline, post-treatment and follow-up (3 months). Effect sizes were large from baseline to post-treatment (MAL-AOU: 9.57; MAL-QOM: 3.24), and from baseline to 3-month follow-up (MAL-AOU: 7.59; MAL-QOM: 1.99).

Taub et al. (1999) reported on ES of the MAL in a sample of patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who received CIMT and reported a large effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
for lower-functioning individuals (n = 11, d = 4.0) and higher functioning individuals (n = 40, d = 3.3). The ES was larger for lower-functioning patients due to lower variability in scores from baseline to post-treatment.

Miltner et al. (1999) reported on ES of the MAL in a sample of 15 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants received two weeks of CIMT and measures were taken at baseline, post-treatment and follow-up (4 weeks and 6 months). Effect sizes were large from first contact to post-treatment (MAL-AOU: 2.07; MAL-QOM: 1.33), from first contact to 4 weeks post-treatment (MAL-AOU: 2.98; MAL-QOM: 1.70), and from first contact to 6-month follow-up (MAL-AOU: 2.68; MAL-QOM: 2.14).

van der Lee et al. (1999) reported on ES of the MAL in a sample of 66 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants were randomly assigned to receive forced manual therapy or bimanual training based on neurodevelopmental techniques for two weeks. A 25-item modified version of the MAL was used. There were no significant between-group differences in MAL-QOM scores following treatment. There was a significant difference in MAL-AOU scores, in favour of forced use therapy. The mean difference in gain was 0.52 points (95% CI, 0.11-0.93). Improvements exceeded the Minimal Clinically Important Difference of 0.50 within both groups. The treatment effect was clinically relevant for patients with hemineglect.

van der Lee et al. (2004) examined responsivenessThe ability of an instrument to detect clinically important change over time.
and longitudinal construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-14 in a sample of 56 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who were randomized to receive CIMT or bimanual training for a 2-week intervention period. ResponsivenessThe ability of an instrument to detect clinically important change over time.
was measured by responsivenessThe ability of an instrument to detect clinically important change over time.
ratios (RR). Results showed adequate responsivenessThe ability of an instrument to detect clinically important change over time.
for the MAL-AOU and MAL-QOM (RR = 1.9, 2.0 respectively). Longitudinal validityLongitudinal validity is the extent to which changes on one measure will correlate with changes on another measure.
was measured by comparing MAL change scores with the Action Research Arm Test (ARAT) change scores and a global change rating (GCR), using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Change scores between measures were not significant nor highly correlated (MAL-AOU vs. ARAT: r = 0.16, p = 0.23; MAL-QOM vs. ARAT: r = 0.16, p = 0.25; MAL-AOU vs. GCR: r = 0.20, p = 0.15; MAL-QOM vs. GCR: r = 0.22, p = 0.10).

Uswatte et al. (2005b) examined responsivenessThe ability of an instrument to detect clinically important change over time.
of the MAL-14 in a sample of 41 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who received CIMT or time-matched general fitness rehabilitation, and their caregivers. ResponsivenessThe ability of an instrument to detect clinically important change over time.
was measured by responsivenessThe ability of an instrument to detect clinically important change over time.
ratios (RR). Results showed high responsivenessThe ability of an instrument to detect clinically important change over time.
for patient scores (MAL-AOU: 3.2; MAL-QOM: 4.5), and caregiver scores (MAL-AOU: 4.3; MAL-QOM: 3.0).

Uswatte et al. (2005b) examined responsivenessThe ability of an instrument to detect clinically important change over time.
of the MAL-14 in a sample of 27 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who received an automated form of constraint-induced movement therapy (AutoCITE) or general fitness rehabilitation. ResponsivenessThe ability of an instrument to detect clinically important change over time.
was measured by responsivenessThe ability of an instrument to detect clinically important change over time.
ratios; results showed high responsivenessThe ability of an instrument to detect clinically important change over time.
for the MAL-AOU and MAL-QOM (RR = 3.8, 5.0, respectively).

Hammer and Lindmark (2010) examined responsivenessThe ability of an instrument to detect clinically important change over time.
and longitudinal construct validityReflects the ability of an instrument to measure an abstract concept, or construct. For some attributes, no gold standard exists. In the absence of a gold standard , construct validation occurs, where theories about the attribute of interest are formed, and then the extent to which the measure under investigation provides results that are consistent with these theories are assessed.
of the MAL-30 in a sample of 30 patients with subacute strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who were randomized to receive forced use therapy or standard upper extremity rehabilitation. ResponsivenessThe ability of an instrument to detect clinically important change over time.
was measured according to effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
(ES), standard response means (SRM) and responsivenessThe ability of an instrument to detect clinically important change over time.
ratios (RR) from baseline to post-treatment (2 weeks), and from baseline to follow-up (3 months). Effect sizes for the MAL-AOU and MAL-QOM were moderate to large from baseline to post-treatment (MAL-AOU: 0.51; MAL-QOM: 0.54) and from baseline to follow-up (MAL-AOU: 1.02; MAL-QOM: 1.17), indicating sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
to change. Standard response means were large from baseline to post-treatment (MAL-AOU: 1.28; MAL-QOM: 1.03), and from baseline to follow-up (MAL-AOU: 1.14; MAL-QOM: 1.19). The greater SRM compared to ES reflects smaller variability in change scores than baseline scores. ResponsivenessThe ability of an instrument to detect clinically important change over time.
ratios were large from baseline to post-treatment (MAL-AOU: 1.22; MAL-QOM: 1.23) and from baseline to follow-up (MAL-AOU: 2.44; MAL-QOM: 2.69). Longitudinal construct was measured by comparison with the FMA-UE, ARAT, Motor Assessment Scale – Upper Extremity score (MAS-UE), 16-hole peg test (16HPT) and the Grippit ratio of isometric grip strength, using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Correlations with the MAS-UE were significant and adequate from baseline to follow-up (MAL-AOU r = 0.53, MAL-QOM r = 0.47); and with the FMA-UE from baseline to post-treatment (MAL-AOU r = 0.44, MAL-QOM r = 0.67) and from baseline to follow-up (MAL-AOU r = 0.39, MAL-QOM r = 0.43).

Khan et al. (2013) examined responsivenessThe ability of an instrument to detect clinically important change over time.
of the German MAL-30 in a sample of 42 patients with acute to chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., using standard response mean (SRM). Participants were stratified into two groups according to level of arm and hand function using the Chedoke McMaster StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Assessment (CSMA). Measures were taken at baseline, discharge from rehabilitation and 6-month follow-up. Change scores from the lower-function group (CSMA arm and hand score ≤ 6) revealed high responsivenessThe ability of an instrument to detect clinically important change over time.
of the MAL-AOU and MAL-QOM from baseline to discharge (SRM = 0.93, 0.94 respectively) and baseline to follow-up (SRM = 0.95. 0.98 respectively), but poor from discharge to follow-up (SRM = 0.20, 0.42 respectively). Change scores from the high-function group (CSMA arm and hand score > 6) showed high responsivenessThe ability of an instrument to detect clinically important change over time.
of the MAL-AOU and MAL-QOM from baseline to discharge (SRM = 1.43, 1.31 respectively) and from baseline to follow-up (SRM = 1.34, 1.33, respectively), but poor responsivenessThe ability of an instrument to detect clinically important change over time.
from discharge to follow-up (SRM = 0.22, 0.24 respectively). The authors concluded that the MAL is a responsive measure when the intervention period is included in the measured time interval.

Simpson & Eng (2013) conducted a literature review of upper limb assessments commonly used in strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. rehabilitation, including the MAL. In studies that measured outcomes following CIMT, the observed change (i.e. patients’ perceptions of change, effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
) was 1.6-6.2 times larger than measures of functional change such as the ARAT or WMFT. Similarly, assessments which measure perceived function in the individual’s environment require larger percentage changes than laboratory-based performance measures to surpass the measurement error. Minimal Detectable Change for the MAL-AOU ranged from 72.5% to 86.7% (90% and 95% confidence levels).

Taub et al. (2013) reported on effect sizeEffect size (ES) is a name given to a family of indices that measure the magnitude of a treatment effect. Unlike significance tests, these indices are independent of sample size. The ES is generally measured in two ways: as the standardized difference between two means, or as the correlation between the independent variable classification and the individual scores on the dependent variable. This correlation is called the “effect size correlation”.
(ES) of the Lower Functioning MAL (LF-MAL) in a sample of 6 individuals with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who used orthotics/splints and adaptive equipment outside the laboratory over 6 sessions (Phase A), then received mCIMT + neurodevelopmental therapy for 15 consecutive weekdays with continued use of assistive devicesAssistive devices are any piece of equipment that you use to make your daily activities easier to perform.
(Phase B). Effect sizes were calculated from (i) baseline to pre-mCIMT; (ii) pre-mCIMT to post-mCIMT; and (iii) baseline to post-mCIMT and were large at all timepoints (ES = 2.6, 2.1, 3.0, respectively, p < 0.002).

Sterr et al. (2014) reported on treatment effect in a sample of 65 patients with chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain.. Participants received four different CIMT treatment protocols that differed in intensity and use of a constraint. Whole-group analysis showed a significant and large treatment effect from baseline to post-treatment (MAL-AOU: d = 1.19; MAL-QOM: d = 1.38); the treatment effect from post-treatment to 6-month follow-up was small but significant for the MAL-AOU only (d = 0.4). Treatment effect was not significant at 12-month follow-up. There was a significant positive association between training intensity and improvement in MAL-AOU scores.

SensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
& SpecificitySpecificity refers to the probability that a diagnostic technique will indicate a negative test result when the condition is absent (true negative).
:
Chen et al. (2012) examined minimal detectable change (MDC)Minimal Detectable Change (MDC) refers to the minimal amount of change outside of error that reflects true change by a patient between two time points (rather than a variation in measurement). of the MAL. This study used data from the EXCITE trial, in which 222 patients with subacute/chronic strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. who were randomized to receive constraint induced movement therapy (CIMT) for 2 weeks (n = 106) or no treatment (n = 116). MDC with 90% confidence intervals was calculated from pre-post test data from the control group. The MDC of the MAL-AOU was 16.8% (Standard Error of the Mean 7.2%), indicating that a change in amount of use of the affected upper limb greater than 16.8% is required so as to be 90% certain that the change is not due to measurement error. The MDC (90% CI) for the MAL-QOM was 15.4% (SEM 6.6%), indicating higher sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
than the MAL-AOU scale. After treatment, the CIMT group showed an 84.6% increase in MAL-AOU scores and a 72.2% increase in MAL-QOM scores. Both MAL scores exceeded the MDC and were sensitive to change in the context of this intervention.

Simpson, Conroy & Beaver (2015) examined sensitivitySensitivity refers to the probability that a diagnostic technique will detect a particular disease or condition when it does indeed exist in a patient (National Multiple Sclerosis Society). See also “Specificity.”
of the MAL-28 in a sample of 9 patients with strokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain., by comparison with the Fugl-Meyer Assessment, the Wolf Motor Function Test and StrokeAlso called a “brain attack” and happens when brain cells die because of inadequate blood flow. 20% of cases are a hemorrhage in the brain caused by a rupture or leakage from a blood vessel. 80% of cases are also know as a “schemic stroke”, or the formation of a blood clot in a vessel supplying blood to the brain. Iimpact Scale. Measures were taken at baseline, post-treatment and follow-up, and correlations were analysed using Spearman’s correlationThe extent to which two or more variables are associated with one another. A correlation can be positive (as one variable increases, the other also increases – for example height and weight typically represent a positive correlation) or negative (as one variable increases, the other decreases – for example as the cost of gasoline goes higher, the number of miles driven decreases. There are a wide variety of methods for measuring correlation including: intraclass correlation coefficients (ICC), the Pearson product-moment correlation coefficient, and the Spearman rank-order correlation.
coefficient. Changes in MAL-AOU scores were sensitive to changes in SIS physical domain scores (ρ = 0.7342, p < 0.0243). Changes in MAL-QOM scores were sensitive to changes in WMFT Functional Ability scores (ρ = 0.6245, p < 0.0722).

References

• Ashford, S., Slade, M., Malaprade, F., & Turner-Stokes, L. (2008). Evaluation of functional outcome measures for the hemiparetic upper limb: a systematic review. Journal of Rehabilitation Medicine, 40, 787-95.
  DOI: 10.2340/16501977-0276
• Cakar, E., Dincer, U., Zeki, M., Kilac, H., Tongur, N., & Taub, E. (2010). Turkish adaptation of Motor Activity Log-28. Turkish Journal of Physical Medicine and Rehabilitation, 56, 1-5.
  http://www.ftrdergisi.com/eng/arsiv.asp
• Chen, H.L., Lin, K.C., Hsieh, Y.W., Wu, C.Y., Liing, R.J., & Chen, C.L. (2018). A study of predictive validity, responsiveness, and minimal clinically important difference of arm accelerometer in real-world activity of patients with chronic stroke. Clinical Rehabilitation, 32(1), 75-83.
  DOI: 10.1177/0269215517712042
• Chen, S., Wolf, S.L., Zhang, Q., Thompson, P.A., & Winstein, C.J. (2012). Minimal detectable change of the Actual Amount of Use Test and the Motor Activity Log: the EXCITE trial. Neurorehabilitation and Neural Repair, 26(5), 507-14.
  DOI: 10.1177/1545968311425048
• Chuang, I.-C., Lin