Motor Assessment Scale

Abstract

The Motor Assessment Scale (MAS) is an objective, performance-based clinical instrument developed in 1985 by Australian physiotherapists Janet H. Carr and Roberta B. Shepherd, alongside psychometricians Lena A. Nordholm and Denise Lynne. Designed to evaluate functional motor recovery following stroke (cerebrovascular accident) and other acquired adult brain injuries, the instrument reflects a paradigm shift away from traditional neurophysiological reflex models toward dynamic systems theory and task-oriented motor learning. The MAS comprises nine distinct functional domains: eight ordinal categories assessing motor behaviors along a hierarchically ordered Guttman-like scale (scored 0 to 6) and one supplementary category evaluating muscle tone (General Tonus). The assessed domains encompass foundational mobility tasks—including Supine to Side-lying, Supine to Sitting over side of bed, Balance Sitting, Sitting to Standing, and Walking—as well as isolated and coordinated upper extremity behaviors: Upper Arm Function, Hand Movements, and Advanced Hand Activities. Psychometric investigations have confirmed the scale’s outstanding reliability and validity across acute, subacute, and chronic rehabilitation settings. Inter-rater reliability coefficients consistently yield Pearson correlation coefficients and Intraclass Correlation Coefficients (ICC) ranging between .89 and .99, while test-retest reliability across repeated sessions demonstrates coefficients exceeding .95. Construct and concurrent validity are strongly supported by significant correlations with established neurorehabilitation metrics, including the Fugl-Meyer Assessment (FMA), the Barthel Index, and the Action Research Arm Test (ARAT). Rasch measurement models and factor analytic studies substantiate that the functional mobility and upper limb subscales demonstrate robust structural validity, distinguishing gross biomechanical transfers from fine distal dexterity, while providing clinicians and researchers with an efficient, highly sensitive metric of motor recovery.

Keywords

Motor Assessment Scale, stroke rehabilitation, motor recovery, neurorehabilitation, task-oriented approach, psychometrics, Guttman scaling, upper limb function, functional mobility, hemiplegia

Authors

The Motor Assessment Scale was formulated and psychometrically evaluated by a multidisciplinary clinical research team based at Cumberland College of Health Sciences (now the Faculty of Medicine and Health, The University of Sydney, Australia):

• Janet H. Carr, EdD, FACP (1933–2014): Pioneering physiotherapist, academic researcher, and educator who served as an Associate Professor in the School of Physiotherapy at the University of Sydney. Carr co-developed the Motor Relearning Programme (MRP) for stroke rehabilitation, which transformed modern neurological physical therapy.
• Roberta B. Shepherd, EdD, FACP: Renowned physiotherapist, clinical scholar, and Emeritus Professor at the University of Sydney. Alongside Janet Carr, Shepherd formulated the biomechanical, task-oriented framework emphasizing ecological dynamics and motor relearning.
• Lena A. Nordholm, PhD: Behavioral scientist and psychometrician who contributed expertise in measurement theory, scale construction, and statistical validation of clinical assessment protocols.
• Denise Lynne, BAppSc (Phty): Clinical physiotherapist who participated in empirical data collection, operationalization of scoring criteria, and inter-rater reliability trials among stroke inpatients.

Purpose

The primary clinical and research objective of the Motor Assessment Scale (MAS) is to provide an objective, task-oriented, time-efficient, and functionally meaningful measurement of motor recovery in individuals presenting with motor dysfunction secondary to upper motor neuron lesions, most notably post-stroke hemiparesis. Historically, stroke evaluation relied heavily on qualitative assessments of muscle tone, primitive reflexes, and isolated joint range of motion (e.g., the Bobath approach or Brunnstrom staging). While informative regarding neuromotor status, these classic systems frequently failed to capture practical everyday motor function, were susceptible to high observer subjectivity, and offered limited predictive power regarding an individual’s capacity to regain independent biomechanical mobility.

Carr and Shepherd established the MAS to operationalize functional performance through directly observable, standardized motor tasks. Clinically, the instrument serves several pivotal functions:

• Baseline Functional Diagnostic Staging: Quantifying the degree of motor impairment across essential daily activities—ranging from bed mobility and trunk stabilization to dynamic balance, ambulation, and fine manual dexterity.
• Treatment Planning and Goal Setting: Identifying the exact biomechanical failure points within an action sequence (e.g., inadequate pelvic anterior tilt during sit-to-stand transfers, or inability to sustain shoulder external rotation during reaching), enabling therapists to structure targeted task-specific exercise regimens.
• Longitudinal Outcome Monitoring: Detecting clinically meaningful changes in motor competence throughout acute, subacute, and community-based rehabilitation phases.
• Distinction Between Compensatory Strategies and Motor Restoration: Because higher tiers within each section mandate specific biomechanical parameters (such as symmetrical weight distribution and absence of excessive trunk flexion or upper extremity synkineses), the MAS differentiates true functional motor recovery from maladaptive compensation.

In empirical research, the MAS functions as an authoritative primary or secondary outcome measure for clinical trials investigating neuroplastic interventions, such as constraint-induced movement therapy (CIMT), robotic-assisted gait training, functional electrical stimulation (FES), and pharmacotherapeutic neurorecovery agents.

Psychological Construct

The latent construct quantified by the MAS is functional motor competence—the capacity of the nervous system to coordinate, execute, and adapt goal-directed musculoskeletal actions within environmental contexts. Rather than treating motor capacity as an abstract, isolated neurophysiological phenomenon, the construct conceptualizes human movement as an emergent, interactive system integrating muscular strength, active range of motion, postural control, sensory feedback, and selective motor control.

Construct Dimensions and Subscales

The instrument parses this overarching construct into eight primary hierarchical motor domains and one diagnostic tone domain:

• 1. Supine to Side-lying onto Intact Side: Measures axial rotation, trunk dissociation, and coordinated limb movement. It reflects early bed mobility, progressing from passive/compensatory rolling using the intact limbs to independent, isolated dynamic initiation by the impaired shoulder and pelvic girdles.
• 2. Supine to Sitting over Side of Bed: Assesses complex multi-segmental transitional mechanics, involving lateral head righting, selective activation of lateral abdominal and trunk musculature, and lower extremity control as gravity assists or resists transition.
• 3. Balance Sitting: Quantifies static and dynamic core postural stability. It moves from externally supported sitting to independent postural equilibrium against self-initiated perturbations, reaching beyond base of support, and selective spinal rotation.
• 4. Sitting to Standing: Evaluates closed-chain lower extremity extension, forward momentum generation, pelvic tilt, and bilateral weight-bearing symmetry. Progression reflects the elimination of manual upper limb weight assistance, achieving full hip/knee extension, and rapid, repeated functional standing cycles.
• 5. Walking: Captures dynamic gait mechanics, temporal-spatial symmetry, and stability. Items advance from unilateral weight-bearing on the paretic leg with assisted swing phase, to unassisted level overground walking, picking up an object from the floor, and navigating stairs independently.
• 6. Upper Arm Function: Measures active proximal motor control of the shoulder complex and elbow. Tasks evaluate isolated protraction, flexion against gravity without synergistic abduction/internal rotation, and overhead stability while suppressing flexor spasticity.
• 7. Hand Movements: Evaluates intermediate motor control, focused on wrist extension, radial deviation, forearm pronation/supination, and gross grasp/release mechanics (e.g., lifting a cylindrical object or grasping a large ball without finger clawing).
• 8. Advanced Hand Activities: Quantifies fine distal motor coordination, independent finger movement, intrinsic hand muscle modulation, and task speed. Activities include pencil manipulation, rapid repetitive tapping (diadochokinesia), spoon usage without compensatory head forward movement, and hair combing.
• 9. General Tonus: A categorical classification indexing resting resistance to passive movement. It tracks deviations from normal compliance, identifying flaccidity, fluctuating tone, or sustained spastic hypertonicity.

Theoretical Framework

The Motor Assessment Scale is grounded in the Motor Relearning Programme (MRP), a framework formulated by Carr and Shepherd that synthesized dynamic systems theory, modern neuroplasticity, and ecological psychology. Prior to Carr and Shepherd’s work, mainstream neurological physical therapy was dominated by hierarchical neuromaturational models, primarily articulated by Signe Brunnstrom, Berta Bobath, and Karel Bobath. These traditional frameworks postulated that motor control develops strictly cranio-caudally and proximo-distally, asserting that recovery following stroke necessitates moving through stereotypic, reflex-dominated synergy stages.

Carr and Shepherd rejected the premise that primitive spastic synergies must be facilitated or elicited as prerequisites for recovery. Instead, drawing upon the pioneering work of Soviet biomechanist Nikolai Bernstein regarding degrees of freedom, and Paul Fitts and Michael Posner’s stages of motor learning, Carr and Shepherd proposed that:

• Movement is goal-directed and task-specific; motor control is organized around functional tasks rather than isolated muscles or stereotypical reflex circuits.
• Impairments following stroke reflect not merely neurological “release phenomena” (such as spasticity), but primary muscle weakness, loss of selective motor activation, secondary biomechanical adaptations (e.g., muscle stiffness and contracture), and learned non-use.
• Motor recovery is an active problem-solving process that relies heavily on sensory feedback, practice, environmental context, and the cognitive engagement of the learner.

The MAS was constructed to mirror this theoretical stance. Tasks within each section are not abstract tests of tendon reflexes or isolated joint movements; they are naturalistic motor behaviors (e.g., picking up a pen top, drinking with a spoon, rolling onto the intact side). The scale enforces Guttman-like hierarchical criteria where success demands optimal biomechanical alignment, minimizing excessive compensatory maneuvers and promoting normal kinematic organization.

Validity

Extensive psychometric investigations have established robust construct, concurrent, predictive, and discriminant validity for the MAS across diverse international clinical cohorts.

Construct and Structural Validity

Construct validity was initially verified by Carr et al. (1985), demonstrating that individuals at varying stages of post-stroke recovery scored differentially across the scale’s sections in a pattern aligning with known physiological recovery trajectories. Proximal trunk and bed mobility tasks generally demonstrate lower difficulty thresholds, whereas complex distal manipulation items (Advanced Hand Activities) present the highest difficulty thresholds.

Subsequent Rasch analysis evaluations (e.g., Poole & Whitney, 1988; English et al., 2006) examined whether the hierarchical structure hypothesized by Carr and Shepherd adheres to strict unidimensional metric assumptions. While investigations confirm that the functional mobility items (Sections 1 through 5) and upper extremity items (Sections 6 through 8) exhibit strong internal ordering, minor item-difficulty reversals have been noted between specific intermediary tiers in varied patient populations, prompting some clinicians to evaluate the motor items as distinct subscale constructs (gross mobility versus upper limb dexterity).

Concurrent and Criterion Validity

Concurrent validity has been established through direct comparison with other gold-standard instruments in neurorehabilitation:

• Fugl-Meyer Assessment (FMA): Strong correlations have been repeatedly documented. Malouin et al. (1994) identified a Spearman rank correlation of $r_s = .88$ to $.96$ between total MAS motor scores and the FMA total motor score. Subscale comparisons between the MAS Upper Arm and Hand items and the FMA Upper Extremity score similarly yield correlations exceeding $.85$.
• Barthel Index (BI): Dean and Mackey (1992) demonstrated strong associations between MAS scores and Barthel Index performance ($r = .78$ to $.86$), confirming that gains in MAS motor competence directly parallel independence in basic activities of daily living (ADL).
• Action Research Arm Test (ARAT): Upper extremity items (6, 7, and 8) show high concurrent validity with ARAT scores ($r > .85$), confirming the scale’s precision in tracking functional reach, grasp, and manipulation.

Predictive Validity

Admission MAS scores during acute and early subacute rehabilitation demonstrate exceptional predictive power regarding long-term functional discharge status, independent community ambulation, and length of hospital stay. Higher baseline performance on the Sitting to Standing and Balance Sitting subscales strongly predicts unassisted community ambulation at six months post-stroke.

Reliability

The Motor Assessment Scale exhibits exceptional reliability parameters when administered by trained healthcare professionals (primarily physical and occupational therapists).

Inter-Rater Reliability

In the seminal psychometric investigation by Carr, Shepherd, Nordholm, and Lynne (1985), thirty stroke patients were independently and simultaneously assessed by pairs of trained therapists. Inter-rater agreement was exceptionally high:

• Pearson product-moment correlation coefficients ($r$) for total scores between examiners ranged from $.95$ to $.99$.
• Individual item correlation coefficients across the nine subscales ranged from $.89$ to $.99$, with the highest concordance observed in Sitting to Standing ($r = .99$) and Walking ($r = .98$).
• Even when calculated via weighted kappa ($\kappa_w$) or Intraclass Correlation Coefficients (ICC), agreement across studies consistently remains well above $.85$, reflecting the objective nature of the standardized scoring criteria (e.g., strict distance, repetition, and timing thresholds).

Test-Retest Reliability

Test-retest stability was evaluated across stable stroke cohorts by Dean and Mackey (1992) and Loewen and Anderson (1988). When evaluations are performed within 24 to 48 hours to minimize actual neurological recovery or fatigue effects:

• Intraclass correlation coefficients for the overall motor score range between $.92$ and $.98$.
• Individual items consistently demonstrate test-retest coefficients between $.82$ and $.95$. The Walking and Balance Sitting categories demonstrate high temporal stability, whereas General Tonus exhibits comparatively lower test-retest reliability ($kappa = .60 – .74$), reflecting natural diurnal fluctuations in spasticity, ambient temperature, posture, and emotional arousal.

Internal Consistency

When evaluated across the eight primary motor subscales (excluding General Tonus), the instrument exhibits strong internal consistency, with Cronbach’s alpha values ranging between $.88$ and $.94$. This high degree of covariance confirms that the motor categories are collectively driven by the underlying continuum of motor recovery.

Factor Analysis

Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) investigations into the MAS have explored whether motor recovery post-stroke functions as a strictly unidimensional construct or presents a multidimensional architecture.

Factor Structures

Structural validation studies, including factor analytic evaluations by English et al. (2006) and earlier structural inquiries by Carr and Shepherd, consistently extract two primary latent dimensions underlying the eight motor sections:

• Factor 1: Gross Motor and Axial/Lower Limb Function: Composed of Section 1 (Supine to Side-lying), Section 2 (Supine to Sitting), Section 3 (Balance Sitting), Section 4 (Sitting to Standing), and Section 5 (Walking). Factor loadings for these mobility-related tasks are consistently high, ranging from $.75$ to $.91$, explaining approximately $50% – 60%$ of the shared variance.
• Factor 2: Fine and Upper Extremity Distal Motor Control: Composed of Section 6 (Upper Arm Function), Section 7 (Hand Movements), and Section 8 (Advanced Hand Activities). Factor loadings on this distinct manual dexterity factor range between $.80$ and $.94$, accounting for an additional $15% – 20%$ of the total variance.

Model Fit and Rasch Dimensionality

When subjected to structural equation modeling and Rasch analysis:

• Goodness of Fit: A two-factor structural model yields superior fit indices compared to a strict single-factor model: Root Mean Square Error of Approximation (RMSEA) values typically fall below $.07$, Comparative Fit Index (CFI) exceeds $.95$, and the Tucker-Lewis Index (TLI) surpasses $.93$.
• Rasch Fit Statistics: Studies utilizing Rasch unidimensional modeling generally demonstrate acceptable infit and outfit Mean Square (MnSq) statistics (falling within the acceptable $0.6 – 1.4$ boundary) when the mobility and upper extremity items are evaluated in functional clusters. The General Tonus category consistently shows substantial misfit and multidimensional distortion if pooled directly with the motor items, empirically validating the clinical convention of reporting tone as a discrete clinical indicator rather than summing it into the total motor score.

Instrument / Measurement Tool

The Motor Assessment Scale is an examiner-administered, performance-based clinical battery. The patient is asked to perform specific physical actions, which the examiner observes and grades against standardized criteria.

Key Structural Characteristics

• Test Type: Performance-based observational functional motor scale.
• Target Population: Adult patients experiencing upper motor neuron lesions, hemiplegia, or hemiparesis following stroke, traumatic brain injury, or tumor resection.
• Total Number of Sections: 9 sections total (8 functional motor performance sections and 1 general muscle tone section).
• Items per Section: Each of the 8 motor sections contains 6 hierarchically arranged functional milestones (scored 0 to 6). The General Tonus section contains 6 categorical descriptors.
• Scoring Range:
  • Motor Subtotal (Sections 1–8): Minimum = 0, Maximum = 48 points.
  • General Tonus (Section 9): Scored categorically; a score of 6 indicates normal tone, whereas deviations (flaccidity, hypertonicity) are qualitatively categorized.
  • Total MAS Score: Commonly reported as 0–48 (motor total) or 0–54 (if including tone where normal = 6). Psychometric consensus strongly recommends reporting the 48-point motor score separate from the tonus score.
• Administration Time: Approximately 15 to 30 minutes, depending on the severity of impairment.
• Equipment Required:
  • Firm plinth or standard hospital bed.
  • Standard height chair or examination stool without armrests.
  • Stopwatch.
  • Measured walking walkway (marking 3 meters, 5 meters, and 10 meters) and a flight of 4 stairs.
  • Small cylindrical object (e.g., small jar or wooden dowel).
  • 5-inch rubber or plastic ball.
  • Polystyrene (Styrofoam) or paper cup.
  • Teacup and 8 small beans or jellybeans.
  • Ballpoint pen with removable cap and standard blank paper.
  • Dessert spoon and water/liquid.
  • Standard hair comb.
• Administration and Scoring Protocol:
  • The patient should be rested and tested in a distraction-free, quiet environment.
  • Testing typically begins at level 1 for each section. If the patient easily achieves level 1, the examiner advances upwards.
  • Each level is scored only if the patient satisfies all operational criteria (including spatial constraints, posture, absence of compensations, and strict time cutoffs).
  • If the patient fails to complete any part of a section, they receive a score of zero (0) for that section.
  • The highest successful level completed represents the score for that respective section.
  • Patients may be permitted up to three attempts per item, with the highest score recorded, provided fatigue does not confound performance.

Permissions & Fee and Test Year

The Motor Assessment Scale was initially formulated and published in 1985 by Janet H. Carr, Roberta B. Shepherd, Lena A. Nordholm, and Denise Lynne in the journal Physical Therapy (American Physical Therapy Association). The instrument was placed in the public domain for clinical, therapeutic, and academic research purposes. No licensing fees, commercial purchase agreements, or formal royalties are required to administer the MAS, download its standardized scoring forms, or publish research findings derived from its use.

Investigators, educators, and rehabilitation clinicians utilizing the instrument are expected to maintain the standardized criteria, adhere to the operationalized instructions without ad-hoc alterations, and cite the foundational 1985 publication and subsequent validation literature in scholarly works.

References

• Carr, J. H., Shepherd, R. B., Nordholm, L., & Lynne, D. (1985). Investigation of a new motor assessment scale for stroke patients. Physical Therapy, 65(2), 175–180. https://doi.org/10.1093/ptj/65.2.175
• Dean, C. M., & Mackey, F. M. (1992). Motor assessment scale scores as a measure of rehabilitation outcome following stroke. Australian Journal of Physiotherapy, 38(1), 31–35. https://doi.org/10.1016/S0004-9514(14)60550-9
• English, C. K., Hillier, S. L., Stiller, K. R., & Warden-Flood, A. (2006). The sensitivity of the Motor Assessment Scale and the Fugl-Meyer Assessment to detect change in motor function in early stroke: A comparative study. Clinical Rehabilitation, 20(1), 83–92. https://doi.org/10.1191/0269215506cr924oa
• Loewen, S. C., & Anderson, B. A. (1988). Reliability of the Modified Motor Assessment Scale and the Barthel Index. Physical Therapy, 68(7), 1077–1081. https://doi.org/10.1093/ptj/68.7.1077
• Malouin, F., Pichard, L., Bonneau, C., Durand, A., & Corriveau, H. (1994). Evaluating motor recovery in stroke patients: A comparison of the Motor Assessment Scale and the Fugl-Meyer Assessment. Physical Therapy, 74(4), 335–345. https://doi.org/10.1093/ptj/74.4.335
• Poole, J. L., & Whitney, S. L. (1988). Motor Assessment Scale for stroke patients: Concurrent validity and interrater reliability. Archives of Physical Medicine and Rehabilitation, 69(3 Pt 1), 195–197.

13. Items of the Scale (Questionnaire)