Distinction between motor learning and motor performance

Physical therapist practice centers on helping patients to learn or relearn motor skills. Motor learning is an internal process associated with practice or experiences that results in a relatively permanent change in a person’s ability to perform a motor skill. Because it is an internal process, motor learning cannot be directly measured and must be indirectly assessed through observation of motor behavior. In physical therapy, motor learning can be indirectly evaluated by measuring change in a patient’s performance of a motor task.

Motor performance may be influenced by a number of variables, including motivation and the use of pharmacologic agents. Maturation and practice are factors that may influence both motor performance and motor learning. A common method of separating the permanent effects of maturation and practice is to measure changes across days or weeks instead of years. However, this method is often ineffective when attempting to measure learning in pediatric and elderly populations because maturation can result in significant changes over days or weeks in children and older adults. To separate maturation and practice influences on performance in these populations, comparisons of practice and nonpractice groups are often necessary. For example, several studies have used two group experimental designs to separate performance changes due to maturation and participation in early intervention programs.^1,2

Temporary factors, such as motivation, physical or verbal guidance, fatigue, stress, and boredom during long therapy sessions, may also influence motor performance. To measure motor learning, the effects of temporary factors on performance should be minimized. The most common method used to reduce the temporary effects of variables on performance is to allow a rest interval between the practice and the evaluation session. In physical therapy settings, the effects of temporary factors can be minimized by evaluating a patient’s performance after he or she rests, or by evaluating performance at the beginning of a subsequent therapy session.

Separating the effects of temporary and permanent factors on performance is critical for documentation. As therapists, we often document changes in patient function based on our observations of a patient’s best performance during a therapy session. The patient’s performance during the session, however, may have been influenced by the temporary effects of our therapy (e.g., hands-on guiding or facilitation techniques). Evaluating and documenting the patient’s performance of a practiced skill at the onset of a subsequent therapy session will more accurately reflect the long-term impact of our interventions.

Physical therapy goals often focus on a patient’s ability to function in various settings under a variety of conditions. For example, goals for gait training might include the ability to walk at slow, medium, and fast velocities; on tile, carpet, grass, or snow; or in a crowded or dimly lit hallway. The field of motor learning distinguishes between assessments in practiced and new or differing environments. Evaluation in the same environment used during a practice or therapy session is termed a retention test, whereas evaluation in a different environment than that used during a practice session is termed a transfer test. For example, if a patient practices walking on a tiled surface during therapy, he or she would undergo a retention test when evaluated on tile, and a transfer test when evaluated on carpet. Retention and transfer tests are used for measures of learning. Retention tests measure how well performers learn practiced tasks. Transfer tests measure how well performers generalize learning to perform the task unpracticed in a different environment.

Overview of processes of motor learning

The processes of motor learning extend beyond conquering the mere motor demands of a task. Motor learning involves developing strategies to cope with the complexities of performing a specific task under specific environmental conditions. Learners must be able to solve motor problems (i.e., process information and engage in sensory encoding and memory retrieval processes) and respond to changes in the task and the environment. To optimize learning, therapists must provide patients with practice conditions that encourage (or possibly force) them to engage in problem-solving processes.³ This suggests that patients should be active participants not only in the production of their movements but also in planning their movements. Instead of providing patients with solutions to motor problems, therapists should act as educators and guide patients through learning processes that allow the patients to actively explore and discover solutions to motor problems.³ A summary of the processes of motor learning is given in Box 14–1.

Motor learning theories

Various theories attempt to explain motor learning.^4–6 Table 14–1 provides a brief summary of several popular theories: Adams’ Closed-Looped Theory,⁴ Schmidt’s Schema Theory,⁵ and Newell’s Ecological Theory.⁶ Each of these theories has limitations and varying levels of research support as well as clinical implications. For example, clinical application of Schmidt’s Schema Theory⁵ may consist of practicing a task under a variety of conditions to assist learners in developing a set of rules or schema related to the specific task. Application of Newell’s Ecological Theory⁶ would suggest that practicing a task under variable conditions will assist learners in understanding the relationship between perceptual cues and motor action (i.e., the glass looks heavy and therefore will require more force to lift).

Table 14–1 Summary of Theories of Motor Learning

Stages of learning

According to Fitts and Posner,⁷ the process of motor learning can be divided into three sequential stages: (1) the cognitive stage, (2) the associative stage, and (3) the autonomous stage. During the cognitive stage, the learner focuses on understanding the task, developing strategies to execute the task, and determining ways to evaluate task success. Performance during this stage is often characterized by inaccuracies, slowness, and movements that appear stiff and uncoordinated. Because this stage is characterized by rapidly improving and variable performance, it is thought to require a high degree of attention and other cognitive processes.

My teenager’s first attempts at driving a car provide a classic example of the cognitive stage of learning. While sitting stiffly and tightly gripping the steering wheel, my daughter would intently focus on the road ahead. As the car moved forward and it came time to shift into a higher gear, her eyes would dart back and forth between the road and the gear shift as she slowly attempted to shift. Engaging in a conversation with me or trying to listen to a traffic update on the radio were too taxing at this stage. All of her attention was directed at trying to keep the car on the road and on understanding the relationships amongst the gas pedal, the clutch, and the brake. Basically, driving demanded all of her attention.

Each time we attempt a new motor task (e.g., juggling, knitting, or snowboarding), and often when we perform a well-learned task in an infrequently practiced environment (e.g., driving a car on icy, snowy roads, or skiing down a steeper hill than we are used to), we find ourselves in this cognitive stage of learning. Our patients often display similar processes during therapy sessions. After a shoulder injury, for example, patients may need to relearn motor tasks such as dressing or brushing their hair.

Therapists can actively promote learning during the cognitive stage by facilitating the patient’s understanding of the task and organizing practice in ways that will specially encourage early learning.⁸ Emphasizing the purpose of the activity within a context that is functionally relevant to the individual patient may help the patient to better understand the task.^8,9 Structuring the environment to reduce distractions may help the learner to better attend to learning. Clear, concise instructions should be provided in a manner that does not overwhelm the learner. Demonstrations of how the task should ideally be performed and providing hands-on guidance as needed will assist the patient in developing a cognitive map of the correct performance of the task. The use of feedback and practice schedules to promote learning at the various stages will be discussed later in this chapter.

After the initial cognitive stage, learners enter the associative stage of learning. Here, the goal is to fine-tune a skill. During this stage, the focus is on how to produce the most efficient action. Relative to the cognitive stage, this stage is characterized by slower gains in performance and reduced variability. To continue with the previous example of learning to drive a car, the first year or so of driving represents the associative stage of learning. After time, my daughter learned to smoothly accelerate and decelerate the car at intersections and to smoothly change gears using the gearshift, clutch, and gas pedals. The associative stage is represented in physical therapy when patients practice a skill to increase the safety or efficiency of a task. For example, when a person with a transfemoral amputation is learning to use a lower extremity prosthesis, the slow transition from taking a few uncoordinated steps to walking smoothly across the floor represents the associative stage. In essence, the patient needs practice time to enhance performance of the skill. Therapists can enhance learning in the associative stage by reducing the amount of hands-on guiding or assistance provided to the patient.⁸ The environment should be structured to gradually promote variations in practice conditions and demands.⁸

The autonomous stage of learning is also described as the automatic stage. Relative to the first two stages, performance in this stage requires very little attention and information processing. After several years of driving practice, my teenager’s driving style characterizes the autonomous stage. She is now able to follow the verbal directions provided by a GPS unit, hold a conversation with a friend in the car, drink from a water bottle, and change lanes on the highway all at the same time. In therapy sessions, this autonomous stage is often achieved over the course of an episode of care when a patient progresses from ambulating with an assistive device, to ambulating independently in select closed environments, to ambulating independently in a community setting under dual task demands (e.g., walking while holding a conversation).

Error detection

The capability to detect errors is another process that is thought to develop with learning. Error detection capabilities are thought to require memory of sensory feedback from previously performed actions and are used differently for slow-positioning and fast-timing tasks. In slow-positioning tasks, sensory feedback is used to guide the action to its endpoint. Thus, learners move until feedback from the present action matches the memory of sensory feedback for the desired action. In fast-timing tasks, learners are unable to use sensory feedback to alter an action online, or during an action. In such tasks, sensory feedback is used to detect errors after the action has ended.¹⁰

Reaching for a cup while reading the newspaper is an example of a slow-positioning task. People will often reach toward the general direction of a cup, then use shoulder abduction and adduction until the hand hits the cup. They will then grasp the cup and bring it to their lips. In contrast, trying to catch a glass of juice that is falling off a table is an example of a fast-timing task. A movement such as this is too fast for sensory feedback to be used during the movement. Sensory feedback can be used only after the movement to determine the accuracy of the action—that is, the person is holding the glass or looking at a puddle on the floor.

Motor memories

Memory is an essential aspect of effective motor learning.¹¹ Three distinct memory systems are thought to exist: short-term sensory store (STSS), short-term memory (STM), and long-term memory (LTM).¹¹ In STSS, numerous segments or streams of sensory information entering the system are briefly stored by sensory modality (e.g., visual, tactile, kinesthetic). Information in STSS is not thought to reach a conscious level and is only retained for a matter of milliseconds. Selected information from STSS is selected for further processing in STM. Selection is thought to be based on the relevance and pertinence of the information. Some authors conceptualize STM as a temporary workspace and may term STM as a form of working memory.¹¹ People hold onto information in STM for only as long as they direct their attention to the information.

LTM is memory system that stores information and experiences accumulated over a life time.¹¹ LTM is thought to consist of two basic forms: declarative (explicit) learning and nondeclarative (implicit) learning. Explicit learning is associated with knowledge that can be consciously recalled and stated, such as facts and events. It requires attention, awareness, and reflection.¹² Implicit learning refers to memories that are less accessible to conscious recollection and verbal recall. Several forms of implicit learning are exemplified by fairly passive learning processes in which learners acquire knowledge through exposure to information. The attentional and cognitive demands of such learning processes are limited. Implicit learning includes the following forms of learning: nonassociative, associative, and procedural.^11,12

Nonassociative learning occurs when individuals are repeatedly exposed to a single stimulus.¹¹ Simple forms of nonassociative learning include habituation and sensitization and are often utilized in clinical practice. For example, patients with certain vestibular disorders have been shown to benefit from exercises that center on repeated performance of activities that provoke their dizziness.^13,14 Although the neuromechanisms behind habituation are not fully understood, over time, these patients habituate to the stimulus and experience a reduction in their symptoms.^13,14

Through associative learning, a person learns to predict relationships such as the relationship of one stimulus to another (classical conditioning) or the relationship of a behavior to a consequence (operant conditioning).¹¹ In therapy sessions, if we repeatedly provide a patient with a verbal cue and a physical prompt when performing a transfer, we may see that over time, the patient will begin to associate the verbal and physical cueing and begin to perform the transfer with verbal cues only. In operant conditioning, a patient may learn that behaviors leading to rewards should be repeated and that behaviors that have negative results should be avoided. This may be a factor for our elderly patients who have recently experienced a fall (a negative consequence) in that such patients may choose to decrease their activity in order to decrease their chances of falling again. Fear of falling may thus become an obstacle that must be conquered in our therapy sessions.

Procedural learning relates to learning tasks though intense practice to the point at which the task can be performed without conscious thought or active attention.¹¹ Such learning occurs with repetition across varying environmental conditions and happens gradually over a period of time. Once the learner learns the “rules” to performing the task, the task is executed with very little conscious attention.¹¹ In therapy, achieving such comfort with motor skills is often a desired outcome. For example, when teaching a patient with a spinal cord injury to perform transfers, our long-term goal may relate to achieving a level of automaticity and generalizability that will allow the patient to transfer safely under a variety of plausible conditions.

Many intervention approaches rely on explicit learning techniques. When teaching crutch use on stairs, for example, therapists often use strategies that involve explicit directions: “Up with the right foot; down with the left foot.” When gait-training with a patient, we often give the patient verbal instructions: “Take a longer step on the right” or “Lift your knee higher.” For patients who are able to process verbal information, maintain attention, and concurrently perform motor and cognitive tasks (listening to the therapist while walking), such strategies may be successful and appropriate. Yet for patients who have cognitive, attentional, or language deficits, such explicit learning strategies may pose difficulties. For example, Orrell and coworkers¹⁵ compared explicit and implicit learning strategies for teaching a dynamic balance task and found that the provision of explicit information may actually be detrimental for some patients after a stroke. A study by Boyd and Winstein¹⁶ further found that regardless of whether subjects had sustained a stoke involving the basal ganglia or the sensorimotor cortex, the provision of explicit information had a negative impact on learning and skill retention.

Focusing on actions, not movements

Many motor behaviorists argue that memories for movements focus on task goals.¹⁷ There is little evidence that learners store and retrieve memories for individual segments of an action (e.g., extend the elbow, open the fingers, close the fingers, then grasp an object), without regard for the task goal or the environment. This principle suggests that patients should practice tasks or actions, not individual movements. For example, a child with cerebral palsy may be learning to ride a tricycle during therapy sessions. The therapeutic goal may be to enhance interlimb coordination between her legs. During practice, however, the therapist and child focus on an outcome goal (moving the tricycle forward as fast as possible), and not on the movements required for interlimb coordination.

Learning to exploit biomechanics

Increased consistency in kinematics and coordination also occurs with practice. Learners are taught to discover ways to take advantage of the passive inertia properties of muscles, joints, and limbs.³ With practice, performers demonstrate increased speed and decreased energy costs because they have learned to optimize the peripheral sensory and motor requirements of the task. Therapists must be able to help patients exploit biomechanics to achieve a goal. This is especially important during the associative stage of learning when patients are trying to fine-tune a skill and should be exposed to different variations of the same skill. For example, therapists most often teach a force-control strategy for sit-to-stand transfers. Although this strategy is relatively safe, a momentum strategy is more efficient.⁹ Shumway-Cook and Woollacott⁹ advocate that patients be allowed to explore several strategies for transfers so that they have choices available. When patients seek safety over efficiency, they may choose a force-control strategy, whereas when efficiency is the primary goal, a momentum strategy may be chosen.