Motor Activity Log

Abstract

The Motor Activity Log (MAL) is a standardized, semi-structured interview developed to quantify real-world upper extremity functional performance in individuals following stroke, traumatic brain injury, and other neurological insults affecting the central nervous system. Developed primarily within the research milieu of Constraint-Induced Movement Therapy (CIMT) by Edward Taub and colleagues at the University of Alabama at Birmingham, the instrument addresses a foundational paradox in neurorehabilitation: the divergence between motor capacity (what a patient can execute under structured laboratory or clinical examination conditions) and spontaneous motor performance (what the patient actually does in the unstructured home and community environment). The standard expanded version of the MAL consists of 30 common activities of daily living (ADL) evaluated across two distinct yet highly correlated subscales: the Amount of Use (AOU) scale and the Quality of Movement (QOM) scale. Both subscales utilize a 6-point ordinal rating structure ranging from 0 to 5, with defined operational descriptors and 0.5-increment intermediate anchors. Psychometric evaluations employing classical test theory and Rasch measurement models confirm that the MAL possesses exceptional internal consistency (α ≥ .88 to .95), robust test-retest reliability (intraclass correlation coefficients typically ≥ .90), strong convergent validity against objective wrist-worn actigraphy and performance-based motor batteries, and profound responsiveness to behavioral neurorehabilitation interventions.

Keywords

Motor Activity Log, MAL, Constraint-Induced Movement Therapy, CIMT, stroke rehabilitation, upper extremity, learned non-use, Amount of Use, Quality of Movement, neuroplasticity, activities of daily living, Rasch analysis, psychometrics, hemiparesis, motor recovery.

Authors

The conceptual genesis and iterative psychometric formalization of the Motor Activity Log reflect several decades of behavioral neuroscience and neurorehabilitation research led by Edward Taub, Ph.D., and his colleagues in the Department of Psychology at the University of Alabama at Birmingham (UAB).

• Edward Taub, Ph.D. – University Professor of Psychology and Director of the CI Therapy Research Group at the University of Alabama at Birmingham (Birmingham, AL, USA). Dr. Taub is renowned for elucidating the phenomenon of learned non-use in non-human primates and pioneering Constraint-Induced Movement Therapy in humans.
• Gitendra Uswatte, Ph.D. – Professor of Psychology and Physical Therapy, University of Alabama at Birmingham. Dr. Uswatte led key psychometric refinement studies, including the formal validation of the 28-item and 30-item iterations and objective cross-validation with wearable accelerometry.
• Victor W. Mark, M.D. – Associate Professor of Physical Medicine and Rehabilitation, University of Alabama at Birmingham. Dr. Mark contributed extensively to evaluating the clinical deployment, stroke phenomenology, and cognitive interactions governing MAL responsiveness.
• Jeanine H. van der Lee, M.D., Ph.D. – Clinical Epidemiologist and Rehabilitation Researcher (Netherlands), who spearheaded the development, cultural adaptation, and clinimetric validation of the 26-item Dutch version of the MAL.

Purpose

The primary clinical and empirical purpose of the Motor Activity Log is to establish an ecologically valid, client-centered measurement of the spontaneous, real-world deployment of the paretic upper limb in naturalistic contexts. In neurorehabilitation, traditional functional tests—such as the Fugl-Meyer Assessment (FMA), the Action Research Arm Test (ARAT), and the Wolf Motor Function Test (WMFT)—measure motor capability under optimal, highly coached laboratory conditions. However, clinicians and clinical neuroscientists have long observed that improvements documented within clinic walls frequently fail to translate into spontaneous utilization of the paretic arm during unmonitored home life. This phenomenon demonstrates that capability does not guarantee habitual usage.

The MAL was deliberately engineered to resolve this diagnostic and evaluative void. Structured as a semi-structured interview administered directly to the patient (or in tandem with an informed caregiver proxy), the MAL systematically guides respondents through a retrospective evaluation of their arm usage over a defined recall period—typically the preceding 24 hours, the preceding week, or across intervention cycles. By anchoring clinical inquiries to concrete, ubiquitous unilateral and bilateral motor tasks (e.g., turning on a light switch, opening a refrigerator, eating with utensils, or buttoning clothing), the instrument minimizes general recall bias and circumvents abstract self-appraisal.

Beyond tracking naturalistic recovery trajectories, the MAL fulfills an indispensable therapeutic function within behavioral neurorehabilitation paradigms such as CIMT. In the context of the CIMT “transfer package”—a suite of behavioral techniques intended to bridge laboratory gains into habitual living—the daily administration of the MAL operates as an adherence monitor, a metacognitive prompt, and a behavioral contract. It forces the patient to self-reflect on motor problem-solving in their domestic ecology, identifying tasks where compensatory strategies (such as excessive reliance on the intact limb) persist unnecessarily. Thus, the MAL serves both as an evaluative endpoint in pivotal clinical trials and as an active behavioral driver of functional motor re-education.

Psychological Construct

The Motor Activity Log measures the multidimensional psychological and behavioral construct of real-world habitual motor performance. Rooted in the International Classification of Functioning, Disability and Health (ICF) framework, the MAL captures the “Performance” domain (what an individual actually executes in their current, unstandardized environment) rather than the “Capacity” domain (the highest probable level of functioning within a standardized testing environment).

This overarching construct is operationalized through two interrelated dimensions:

1. Amount of Use (AOU)

The Amount of Use subscale assesses the quantitative frequency and relative engagement of the weaker, paretic extremity during specified activities of daily living relative to the premorbid baseline. It does not measure whether an isolated movement was kinematically flawless; rather, it queries behavioral selection: When the opportunity arose to open a door, pull up trousers, or pick up a glass, was the affected limb recruited, ignored, or substituted by compensatory adaptations? A score of 0 denotes complete behavioral omission (learned non-use), whereas higher scores reflect progressive shifts toward habitual, spontaneous inclusion of the affected limb as a primary or stabilizing manipulator.

2. Quality of Movement (QOM)

The Quality of Movement subscale captures the patient’s subjective and metacognitive appraisal of the motor execution’s biomechanical efficiency, smoothness, precision, and effort. While clinical observers evaluate kinematics via three-dimensional motion capture or observational scales, the patient’s lived experience of movement quality involves internal perceptions of spasticity, ataxia, synergy emergence, slow velocity, and compensatory trunk displacement. Under the QOM framework, the respondent evaluates whether the movement was carried out with severe difficulty, awkwardness, and sluggishness (low scores) versus smooth, rapid, and normal control indistinguishable from pre-stroke functionality (high scores).

The Interplay of Metacognition, Self-Efficacy, and Compensatory Habits

The construct captured by the MAL is deeply psychological. It is mediated by the individual’s perceived self-efficacy regarding limb safety and capability. Following a cerebrovascular accident, attempts to utilize the paretic limb are frequently met with failure, embarrassment, prolonged task completion times, or physical discomfort. This conditioned negative feedback induces a self-reinforcing cognitive schema wherein the individual defaults to the non-paretic limb even when neurological recovery would permit paretic limb deployment. The MAL quantitatively maps this behavioral habit pattern, charting the psychological threshold where deliberate, effortful limb selection transitions back into automatic, effortless motor behavior.

Theoretical Framework

The theoretical cornerstone of the Motor Activity Log is the Learned Non-Use Theory, conceptualized by Edward Taub based on classical deafferentation experiments in non-human primates and subsequent clinical investigations in humans recovering from neurological trauma.

Following focal brain damage, the immediate post-injury epoch is characterized by diaschisis, neural shock, and profound neuromuscular weakness. During this critical window, early attempts by the patient to engage the affected limb inevitably result in motor failure, clumsiness, and frustration. According to principles of operant conditioning, these unsuccessful attempts act as positive punishment or extinction trials, suppressing future behavioral motor initiation. Concurrently, the patient discovers that using the intact, unaffected limb reliably achieves functional objectives (e.g., retrieving objects, eating, personal hygiene), thereby conferring immediate positive and negative reinforcement.

Consequently, the behavioral habit of relying solely on the unimpaired extremity becomes deeply entrenched. Long after diaschisis has resolved and biological healing has restored latent physiological capacity, this conditioned non-use persists indefinitely. At the neurobiological level, this behavioral suppression induces maladaptive cortical reorganization: the cortical motor representation corresponding to the paretic arm within the primary motor cortex shrinks from lack of behavioral drive (use-dependent cortical depression).

The MAL is theoretically grounded in this operant and neuroplastic matrix. In evaluating both the quantity and perceived quality of spontaneous movements, it measures the degree to which an individual remains trapped within the vicious cycle of learned non-use or, conversely, has successfully overridden conditioned suppression through intensive behavioral shaping and forced-use paradigms. The theoretical premise dictates that real-world functional recovery requires expanding cortical motor zones through thousands of task repetitions, a transformation that can only be sustained if real-world behavioral frequency (MAL score) reaches a self-sustaining threshold.

Validity

Extensive psychometric investigations have established the validity of the Motor Activity Log across stroke, cerebral palsy, multiple sclerosis, and traumatic brain injury cohorts.

Construct and Structural Validity

Structural validity has been corroborated through both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA), as well as Rasch modeling. Unidimensionality is strongly supported for both the AOU and QOM subscales independently. Rasch analyses conducted on the MAL-30 and MAL-28 demonstrate that the items conform to a hierarchical continuum of motor difficulty. In studies evaluating stroke survivors (e.g., Uswatte et al., 2005, 2006; van der Lee et al., 2004), item infit and outfit mean square (MnSq) statistics fall predominantly within the acceptable psychometric range of 0.6 to 1.4, confirming that the individual tasks contribute coherently to measuring a single underlying functional construct without substantial multidimensional distortion.

Convergent Validity

Convergent validity has been evaluated by contrasting MAL scores with performance-based laboratory assessments and objective wearable technology:

• Objective Accelerometry: Gitendra Uswatte and colleagues demonstrated that MAL Amount of Use scores correlate moderately to strongly with objective wrist accelerometry ratios (ratio of paretic to non-paretic arm movement counts over multiple days), yielding correlation coefficients ranging from r = .52 to r = .74 (p < .001). This confirms that self-reported MAL scores reliably reflect objective, continuous physical limb activity in the home ecology.
• Laboratory Capacity Tests: The MAL demonstrates significant positive correlations with the Action Research Arm Test (ARAT; r ≈ .60 to .75) and the Wolf Motor Function Test (WMFT; r ≈ −.55 to −.72 with WMFT performance time, and r ≈ .65 to .80 with WMFT functional ability scale). Correlations with the upper extremity Fugl-Meyer Assessment (FMA) are typically moderate (r ≈ .50 to .65), reflecting the conceptual divergence between pure motor impairment (synergistic reflexes) and integrated functional ADL performance.

Discriminant and Known-Groups Validity

The MAL accurately discriminates between varying degrees of stroke severity and levels of arm paresis. It cleanly distinguishes individuals with mild, moderate, and severe hemiparesis. Furthermore, it differentiates patients undergoing conventional physical or occupational therapy from those receiving high-intensity behavioral interventions like CIMT, revealing statistically dramatic group-by-time interactions.

Responsiveness and Minimal Clinically Important Difference (MCID)

The MAL is recognized as one of the most sensitive outcome instruments in neurorehabilitation, consistently exhibiting large effect sizes (Cohen’s d > 1.0; often exceeding 1.5 to 2.0) in CIMT clinical trials. Distribution-based and anchor-based investigations estimate the Minimal Clinically Important Difference (MCID) to be approximately 0.50 points on both the AOU and QOM subscales for chronic stroke populations. Changes of this magnitude reflect a noticeable, meaningful transition in everyday living—such as moving from entirely passive or assisted limb usage to spontaneous, purposeful bimanual integration.

Reliability

The Motor Activity Log exhibits exceptional reliability metrics across classical psychometric paradigms, establishing its consistency across test administrations, raters, and clinical environments.

Internal Consistency

Internal consistency estimates for the MAL are consistently high across diverse clinical cohorts:

• Amount of Use (AOU): Cronbach’s alpha coefficients consistently range between α = .88 and .96.
• Quality of Movement (QOM): Cronbach’s alpha coefficients similarly demonstrate high internal consistency, ranging from α = .90 to .97.
• Rasch item reliability indices exceed .95, and person reliability indices typically exceed .85, confirming that the selected 30 items thoroughly sample the functional continuum with minimal measurement redundancy.

Test-Retest Reliability

Given that the MAL is an evaluative interview reflecting habitual behaviors, test-retest stability is critical. In chronic, clinically stable stroke populations evaluated across intervals ranging from 3 to 14 days, the Intraclass Correlation Coefficients (ICC) are exceptionally high:

• ICC values for the AOU scale range between .91 and .98.
• ICC values for the QOM scale range between .89 and .96.

These values demonstrate that when functional status is stable, self-report variance does not introduce meaningful measurement error.

Inter-Rater and Proxy Agreement

Because the MAL is administered as a structured interview, inter-rater reliability assesses whether different trained clinicians extract identical ratings from the same individual. Studies utilizing standardized administration protocols report inter-rater ICCs exceeding .90. Furthermore, studies comparing patient self-reports with ratings provided independently by living-in caregivers indicate moderate-to-strong agreement (r = .65 to .82), demonstrating that the structured anchoring cues effectively constrain subjective inflation.

Factor Analysis

Extensive factor-analytic investigations have scrutinized the latent architecture of the Motor Activity Log, particularly regarding whether the AOU and QOM scales capture distinct dimensions or reflect a singular underlying functional construct.

Dimensionality and Collinearity

In classical factor-analytic studies, EFA consistently reveals that when AOU and QOM items are analyzed separately, each subscale forms a distinct, dominant first factor accounting for upwards of 55% to 68% of the total variance. Factor loadings for individual items across both subscales typically exceed .60, with the vast majority falling between .70 and .88 (e.g., tasks such as “open a refrigerator door,” “eat with a fork or spoon,” and “pick up a telephone” exhibit exceptionally high loadings).

However, when AOU and QOM are subjected to simultaneous joint factor extraction, a profound psychometric phenomenon emerges: the two scales correlate extremely highly with one another (often r ≥ .85 to .93). CFA models testing a two-factor correlated model versus a single general factor model frequently reveal that while a two-factor model yields slightly superior goodness-of-fit indices (e.g., CFI > .94, TLI > .93, RMSEA ≈ .06), the high inter-factor correlation indicates that in the patient’s subjective appraisal, how often the arm is used is inseparable from how well it moves.

Rasch Measurement Model Results

To further examine the scale’s measurement structure, Gitendra Uswatte, Edward Taub, and colleagues (2005, 2006) conducted item-response and Rasch analyses on the 30-item and 28-item MAL. Their findings established that:

1. The rating scale categories progress monotonically; the 6-point operational anchors (and half-point increments) function predictably without disordered thresholds, demonstrating that respondents differentiate between the incremental levels of functional return.
2. The item difficulty hierarchy aligns with clinical biomechanics. Tasks with low stabilization demands (e.g., Item 1: “Turn on a light with a light switch,” Item 5: “Wipe off a kitchen counter”) represent easy items, anchoring the lower end of the logit scale. Conversely, tasks demanding fine distal motor control, precise prehension, and fractionated finger movement (e.g., Item 25: “Key / unlock a door with a key,” Item 26: “Write with a pen or pencil,” Item 14: “Tie shoes”) represent the most difficult items, anchoring the upper end of the logit scale.
3. Differential Item Functioning (DIF) analyses indicate that the item hierarchy remains stable across age cohorts, sex, and side of hemiparesis (dominant vs. non-dominant hemisphere involvement).

Instrument / Measurement Tool

The Motor Activity Log is structured as a standardized semi-structured interview administered by a trained health professional (physical therapist, occupational therapist, or clinical researcher).

• Test Type: Patient-Reported Outcome Measure (PROM) / Clinician-Administered Semi-Structured Interview.
• Target Population: Adult and geriatric patients presenting with upper extremity paresis resulting from stroke, traumatic brain injury, central nervous system lesions, or cerebral palsy.
• Administration Format: Face-to-face or telehealth semi-structured interview, utilizing laminated visual response cards displayed before the patient.
• Completion Time: Approximately 20 to 30 minutes for the full 30-item evaluation.
• Item Count: 30 core activities of daily living (variations include the 14-item abbreviated scale, the 28-item validated iteration, and the 26-item Dutch translation).
• Response Scales: Two distinct 6-point ordinal scales with half-step increments permitted (11 distinct scoring levels: 0.0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0):
• Amount of Use (AOU) Scale:
  • 0: Did not use weaker arm.
  • 1: Very rarely used weaker arm.
  • 2: Rarely used weaker arm.
  • 3: Half as much as before stroke/injury.
  • 4: Almost as much as before stroke/injury.
  • 5: As much as before stroke/injury (normal use).
• Quality of Movement (QOM) Scale:
  • 0: Weaker arm was not used at all.
  • 1: Weaker arm was used, but movements were extremely slow, awkward, and with very great difficulty.
  • 2: Weaker arm was used, but movements were slow, clumsy, and accompanied by great difficulty.
  • 3: Weaker arm was used, but movements were sluggish and accompanied by moderate difficulty.
  • 4: Weaker arm was used almost normally, with only slight slowness or clumsiness.
  • 5: Weaker arm moved normally/as well as before stroke/injury.
• Administration Guidelines:
  • The interviewer presents a visual scale card to the respondent displaying the verbal descriptors for each numerical level.
  • For each item, the interviewer reads the activity description verbatim. The clinician asks whether the patient completed the activity during the specified recall interval (e.g., past week).
  • If the activity was performed, the interviewer queries: (1) How much was the weaker arm used? and (2) How well did the weaker arm move during that activity?
  • If an activity was not completed during the recall period for environmental or incidental reasons unrelated to arm disability (e.g., the patient does not own a computer, hence did not type), that item is designated as Not Applicable (N/A) and omitted from the denominator.
• Scoring and Computational Rules:
  • The MAL does not utilize a cumulative sum score.
  • The total Amount of Use (AOU) score is calculated as the mathematical mean of all valid scored items:
    
    Mean AOU = (Σ AOU Item Scores) / (Total Number of Applicable Items Scored)
  • The total Quality of Movement (QOM) score is calculated similarly:
    
    Mean QOM = (Σ QOM Item Scores) / (Total Number of Applicable Items Scored)
  • Overall scores range continuously from 0.00 to 5.00.
  • There is no reverse scoring; all items are coded in a positive functional direction.

Permissions & Fee and Test Year

The Motor Activity Log was originally introduced in its preliminary 14-item form by Edward Taub and colleagues in 1993, followed by extensive clinimetric standardizations and expansions into the 30-item format (Uswatte et al., 2005) and 28-item psychometric version (Uswatte et al., 2006). A Dutch adaptation of 26 items was standardized by van der Lee and colleagues in 2004.

Copyright and Licensing: The Motor Activity Log is non-commercial and resides within the public academic and clinical domain for research and clinical rehabilitation applications. It was developed primarily through grants funded by the National Institutes of Health (NIH) awarded to the University of Alabama at Birmingham. Clinicians and scientists may utilize the instrument without royalty fees. However, researchers are expected to preserve the structural integrity of the instructions, operational anchors, and item wordings, and to formally cite the primary validation manuscripts established by the UAB CI Therapy Research Group. Standardized interview administration manuals, scoring sheets, and caregiver versions are disseminated openly through university research repositories and clinical trial documentation networks.

References

• Taub, E., Miller, N. E., Novack, T. A., Cook, E. W., Fleming, W. C., Nepomuceno, C. S., Connell, J. S., & Crago, J. E. (1993). Technique to improve chronic motor deficit after stroke. Archives of Physical Medicine and Rehabilitation, 74(4), 347–354. https://doi.org/10.1016/0003-9993(93)90004-k
• Taub, E., Crago, J. E., & Uswatte, G. (1998). Constraint-induced movement therapy: A new approach to neurorehabilitation in stroke patients. Physical Medicine and Rehabilitation Clinics of North America, 9(2), 527–557. https://doi.org/10.1016/S1047-9651(18)30282-3
• Taub, E., Uswatte, G., & Pidikiti, R. (1999). Constraint-Induced Movement Therapy: A new family of techniques with broad application to physical rehabilitation–a clinical review. Journal of Rehabilitation Research and Development, 36(3), 237–251.
• Uswatte, G., Taub, E., Morris, D., Vignolo, M., & Lin, K. (2005). Reliability and validity of the upper-extremity Motor Activity Log-14 for measuring real-world arm use. Stroke, 36(11), 2493–2496. https://doi.org/10.1161/01.STR.0000185928.90848.2e
• Uswatte, G., Taub, E., Morris, D., Temple, E., & Vignolo, M. (2006). Contribution of the expanded Motor Activity Log to the assessment of real-world arm use after stroke: The MAL-28 and MAL-30. Physical Therapy, 86(10), 1384–1393. https://doi.org/10.2522/ptj.20050186
• van der Lee, J. H., Beckerman, H., Knol, D. L., de Vet, H. C., & Bouter, L. M. (2004). Clinimetric properties of the Motor Activity Log for the assessment of arm use in hemiparetic patients. Stroke, 35(6), 1410–1414. https://doi.org/10.1161/01.STR.0000126819.82090.C2
• Wolf, S. L., Winstein, C. J., Miller, J. P., Taub, E., Uswatte, G., Morris, D., Giuliani, C., Light, K. E., & Nichols-Larsen, D. (2006). Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: The EXCITE randomized clinical trial. JAMA, 296(17), 2095–2104. https://doi.org/10.1001/jama.296.17.2095

13. Items of the Scale (Questionnaire)