Book summary “Motor learning and performance”: summary

Introduction

You face a difficult challenge when you try to master a complex physical skill in sports, physical therapy, or everyday life. You invest countless hours in practice, yet your performance often drops during actual competition or testing.

Richard A. Schmidt provides a comprehensive framework to solve this problem in Motor learning and performance. The central idea is that mastering motor skills requires navigating information processing limits, managing feedback schedules, and structuring practice conditions to optimize permanent capability gains.

The book divides this study into two parts. You first explore human skilled performance to understand how the nervous system plans and executes rapid movements. You then examine skill learning principles to discover how to design effective practice environments.

This text answers several core questions about human movement:

• How do you process sensory information during high-speed actions?
• Why does random practice improve long-term retention despite causing poor immediate performance?
• What underlying abilities determine your ultimate potential in a given motor skill?

1. Defining Skills and Information Processing

Motor skills require achieving environmental goals with maximum certainty and minimum outlay of time and energy.

Richard A. Schmidt defines a skill through three core pillars. You maximize the certainty of goal achievement while minimizing physical and mental energy costs. You also minimize the time used.

Gymnast Gabby Douglas winning the Olympic all-around gold medal in 2012 demonstrates the pinnacle of a highly practiced skill executed with maximum certainty.

Classification of Skills

You can categorize these actions along several distinct axes.

• Open skill: occurs in a variable and unpredictable environment requiring constant adaptation
• Closed skill: occurs in a stable and predictable environment allowing advance organization
• Discrete skill: has a clearly defined beginning and end and a brief duration
• Continuous skill: has no specific beginning or end and flows continuously for minutes
• Serial skill: consists of multiple discrete actions combined in a specific order

Driving a car in heavy traffic represents an open skill, whereas swimming in an empty pool lane is a closed skill. Firing a rifle is a discrete skill, steering a car along a road is a continuous tracking task, and performing a gymnastics routine is a serial skill.

Information Processing Stages

Executing any skill requires first perceiving relevant environmental features. Next, you must decide what to do and when. Finally, the motor system must produce organized muscular activity to generate the required movement, which includes postural and body transport components.

A surgeon must perceive a delicate situation, decide on a precise course of action, and execute movements supported by a stable postural platform.

Human motor performance relies on three sequential, nonoverlapping stages of information processing:

• Stimulus identification: detects and assembles environmental stimuli into a representation
• Response selection: decides what response to make based on the environment
• Movement programming: organizes the motor system and retrieves the motor program

Driving a car involves seeing a hazard in traffic, deciding whether to brake or swerve, and finally programming the physical movement to press the brake pedal.

Reaction Time and Decision Making

Reaction time measures the speed of decision making. Reaction time is the time elapsing from an unanticipated stimulus until the beginning of a response. Movement time is the period from the end of reaction time until movement completion. Response time is the total sum of reaction time and movement time combined.

When pressing a car brake, the time to lift the foot from the accelerator and depress the pedal includes both the reaction time and the movement time.

Hick’s Law states that choice reaction time increases linearly as a function of the logarithm of the number of stimulus-response alternatives. Doubling the number of alternatives increases choice reaction time by a constant amount.

Merkel's 1885 experiment asked subjects to press keys for up to 10 possible Arabic and Roman numeral stimuli, showing a sharp rise in reaction time from 1 to 2 alternatives and smaller increments toward 10 alternatives.

Stimulus–response compatibility refers to the naturalness of the connection between a stimulus and a response. Spatially compatible arrangements lead to a faster resolution of uncertainty in the response selection stage and shorter choice reaction times compared to incompatible arrangements.

Turning bicycle handlebars to the right to move right is directionally compatible, whereas steering a sailboat requires moving the tiller left to turn right.

2. Attention Capacity and Sensory Control

Attentional resources are strictly limited, and movement control relies on integrating exteroceptive and proprioceptive sensory feedback.

Attention and Performance Limits

Attention functions as a limited capacity resource or bank account that must be allocated among competing tasks. When two tasks compete for limited attentional capacity, they interfere with each other. A complex primary task leaves less attention available for secondary tasks.

Driving a car while trying to remember a forgotten shopping list, answering a phone call, monitoring traffic lights, and adjusting climate controls illustrates this resource competition.

Controlled processing is slow, effortful, serial, and attention-demanding. Automatic processing resulting from extensive practice is fast, parallel, and requires little attention. Controlled processing governs novel or poorly learned activities that interfere with concurrent tasks. Automatic processing allows highly skilled performers to execute routines quickly and in parallel without disrupting performance, though it can cause errors if the environment changes unexpectedly.

Olympic gymnast Peter Vidmar paid attention only to the first move of his routine, while the rest occurred automatically. This allowed him to focus on style and form during the performance.

Sensory Contributions and Feedback Loops

Sensory information for motor performance is categorized into exteroceptive data from the environment and proprioceptive data from within the body. Exteroceptive feedback provides information about the state of the surrounding environment, primarily through vision and hearing. Proprioceptive feedback supplies information about the state of the body itself, including joint positions, muscle tensions, and balance, originating from specialized receptors.

Human motor control can be modeled as a closed-loop system where actual output is compared against a reference to detect and correct errors. A basic closed-loop control system relies on feedback to regulate performance toward a goal.

Maintaining room temperature using a thermostat and a furnace, or reaching out to pick up a cup while visually tracking hand position, demonstrates this comparative regulation. Closed-loop control models are effective for slow or continuous tasks but fail to account for rapid, discrete movements due to processing sluggishness. Tracking systems can only produce corrections at a maximum rate of about three per second. For rapid actions lasting less than 300 ms, feedback cannot be processed in time to alter the ongoing movement.

Visual Control Systems

Human vision operates via two separate processing streams: the ventral stream for conscious object identification and the dorsal stream for nonconscious movement control. Visual information from the retina splits at the primary visual cortex. The ventral stream sends data to the inferotemporal cortex to answer what an object is. The dorsal stream sends data to the posterior parietal cortex to answer where an object is across the entire visual field.

Neurological patients TN and DB demonstrated blindsight by successfully navigating a hallway or guessing object locations without conscious sight. Optical flow across the retina provides the dorsal visual system with real-time environmental information regarding time-to-contact, direction, and balance. The optical variable tau, defined as retinal image size divided by its rate of change in size, is proportional to the time remaining until contact.

David N. Lee used a moving room apparatus where shifting surrounding walls caused young children to sway forward or sit down based on false optical flow.

3. Motor Programs and Speed Accuracy Trade Offs

An open-loop control system consists of an executive and an effector, lacking feedback and comparator mechanisms, making it effective for predictable environments but inflexible to sudden changes. Traffic signals and many microwave ovens operate on an open-loop system, following programmed timing without reacting to unexpected real-time events.

Rapid actions are governed by generalized motor programs in an open loop fashion, subject to physiological force variability and speed accuracy trade offs.

Evidence for Motor Programs

Reaction time increases as the complexity of the movement to be performed increases, indicating that actions are organized in advance during the movement programming stage. Henry and Rogers (1960) found that simple reaction time increased from 150 ms for a simple finger lift to 195 ms for an intermediate reach-and-slap movement, and 208 ms for a complex multi-step movement with two directional reversals.

Generalized Motor Programs

Generalized Motor Program theory solves the storage and novelty problems by proposing that a single stored program represents a class of actions defined by invariant features, while surface features are adjusted via parameters. A person can write their signature or a palindrome using their dominant hand, immobilized wrist, nondominant hand, teeth, or foot, while preserving the underlying movement structure.

Speed and Accuracy Principles

Movement time increases systematically as movement amplitude increases or target width decreases, following a logarithmic relationship known as Fitts' Law. Fitts used tasks where subjects tapped alternately between two target plates, transferred washers between pegs, or moved styluses across varied widths and amplitudes.

In rapid open-loop movements, movement inaccuracy increases linearly as average movement velocity increases. Schmidt and colleagues tested rapid aiming movements with movement times of 200 ms or less, plotting variability of end points against average movement velocity.

Rapid movements become spatially inaccurate because stronger muscle contractions generate higher force variability, causing the limb to deviate from its intended trajectory. Sherwood, Schmidt, and Walter studied brief ballistic force applications to a handle, showing that force variability increases with the percentage of maximum force used up to 70 percent of maximum force.

For skills where temporal accuracy is the primary goal, reducing movement time decreases timing error and makes the action more accurate in time. Schmidt, Zelaznik, and Frank demonstrated that halving the goal movement time halves timing errors in rapid temporal tasks from 300 ms to 150 ms goal movement time reductions.

You can test these timing principles yourself:

• Use the timer function on a cell phone or watch without looking at it.
• Start and stop the timer to generate exactly 10 seconds and record the error over 10 trials.
• Repeat the 10 trials trying to generate 5 seconds and compare the error measures.

When two limbs are assigned separate or incongruent spatial goals, performance departs from independent Fitts' Law predictions due to a joint command structure. Kelso and colleagues used bimanual Fitts tasks, barrier-clearing studies, and the gamma-V experiment where subjects simultaneously drew a 'V' with one hand and a 'γ' with the other in 2 in. (5 cm) figure dimensions for the gamma letter.

In continuous bimanual tasks, rhythmic movements spontaneously adopt an in-phase coordination pattern because it offers superior stability and resistance to high speeds compared to anti-phase patterns. Kelso, Scholz, and Schöner had subjects perform rhythmic finger movements at gradually increasing tempos, observing spontaneous pattern transitions at higher frequencies past a 2.25 Hz frequency threshold where anti-phase patterns switch to in-phase from 180 degrees relative phase for anti-phase patterns.

4. Individual Differences and Abilities

Motor behavior research splits into the experimental method, which assumes people behave uniformly and treats individual variations as error, and the differential method, which focuses on these very differences between people. The experimental method studies how variables affect human motor behavior in general. In contrast, differential psychology investigates stable differences among people using statistical tools like correlations.

Human motor performance depends on independent, inherited abilities that shift in composition as learners transition through practice.

Abilities Versus Skills

An ability serves as basic inherited equipment that underlies various real-world tasks and remains unmodifiable by practice. A skill represents potential and proficiency at a specific activity and can be continually developed and refined.

Decathlete Bryan Clay achieved elite performances in ten diverse events at the 2008 Summer Olympics. Misty May Treanor and Keri Walsh Jennings leveraged their 6 ft 3 in (1.9 m) height abilities to win gold in beach volleyball.

Specificity of Motor Abilities

Franklin Henry's specificity hypothesis states that there is no single general motor ability. Motor abilities are instead highly specific to particular tasks. Statistical correlations computed between different movement skills are generally very low, showing that possessing proficiency in one task does not predict proficiency in another.

Drowatzky and Zuccato examined six balance tests and found the highest correlation coefficient ($r = .31$) between the bass stand and sideward stand balance tests. These tests shared only 9.6 percent of their underlying abilities. Edwin Fleishman used factor analysis to demonstrate that human performance depends on a multitude of separate, independent motor and physical proficiency abilities, estimated between 20 to 50 in number. These include reaction time, response orientation, speed of movement, and response integration.

Predicting Skill Success

The pattern of abilities underlying a particular task changes as a learner progresses from a novice to an expert through practice and experience. Novice performance often requires heavy cognitive and spatial processing abilities that drop out over time.

Fleishman and Hempel studied subjects practicing a complex visual discrimination reaction-time task. They found that the contribution of spatial relations dropped from 30 percent on the first trial to 5 percent by the fifteenth trial. Meanwhile, the final contribution value of specific discrimination reaction time reached 40 percent as the largest single contributor.

5. Organizing and Scheduling Practice

Distributing practice intervals and implementing random practice schedules enhance long term memory retention by increasing contextual interference.

You can structure training sessions to maximize permanent gains rather than temporary performance. The arrangement of rest and task variations determines how well you encode skills into long term memory.

Practice and Rest Intervals

Distributing practice across shorter sessions over more days yields vastly superior long term learning compared to massed, highly concentrated practice schedules. Concentrated practice causes fatigue and limits the consolidation of learning.

Baddeley and Longman tested postal workers on keyboard tasks for 80 total hours. They found that the group practicing 4 hours per day learned the least, despite finding the schedule most appealing compared to the 1 hour per day group who practiced for 60 total hours.

The length of rest intervals between trials strongly impacts continuous motor tasks due to cumulative fatigue. Rapid discrete movements show minimal decrements even with extremely short rest.

Bourne and Archer had subjects perform 30 second trials on a pursuit rotor task separated by rest intervals ranging from 0 to 60 seconds. Longer rest intervals lead to more skilled performance and learning.

Variable Versus Constant Practice

Variable practice enhances the development of schemas, allowing you to generate novel actions and generalize past learning to unexperienced task variations. Practicing multiple members of a class of actions helps you abstract the mathematical relationship between surface parameters and movement outcomes.

Catalano and Kleiner had groups practice a coincident timing task at constant or variable speeds using 5, 7, 9, and 11 mph training velocities. Variable practice led to smaller errors during transfer tests at novel speeds of 1, 3, 13, and 15 mph.

Blocked Versus Random Practice

Random practice schedules produce inferior performance during acquisition but vastly superior retention and learning compared to blocked practice schedules. Interleaving different tasks prevents automatic repetition and forces deeper cognitive processing.

Shea and Morgan had subjects practice three different movement sequence tasks in either a blocked or random order. Random practice groups performed much better on delayed retention test intervals lasting 10 minutes and 10 days.

Random practice benefits learning through two distinct mechanisms:

• The elaboration hypothesis, by creating more distinct and meaningful memories through task comparisons.
• The forgetting hypothesis, by forcing the regeneration of short term solutions on every trial.

Wright showed that blocked practice groups explicitly instructed to compare tasks outperformed standard groups. Lee et al. found that providing a guiding model eliminated the benefits of random practice by removing reconstruction.

Mental rehearsal of skills without overt physical movement generates actual motor learning and improves task performance. Hird and coauthors demonstrated that mental practice produced significant learning gains compared to a control condition on pegboard insertion and pursuit rotor tasks.

You can apply these mental rehearsal strategies during training:

• Rehearse skills to be learned mentally without performing actual physical practice.
• Think about the skills being learned and rehearse each step sequentially.
• Imagine doing the actions that result in achieving the goal.
• Interleave mental practice during rest phases between trials of fatiguing tasks.

6. Augmented Feedback Principles

Feedback in motor skills is split into inherent feedback coming naturally from the action and augmented feedback supplied artificially by an instructor or device. Inherent feedback includes sensory consequences like seeing a ball travel or feeling body movements. Augmented feedback is supplemental information over which an instructor has direct control, such as verbal scores or video replays, and serves as the primary tool for shaping practice organization.

A tennis player feels their arms moving and hears ball contact inherently, but receives augmented feedback when a coach comments on their stroke or a digital screen displays a score. Knowledge of results evaluates action success regarding environmental goals, whereas knowledge of performance details the movement pattern itself. Knowledge of results is usually verbalizable information about success relative to the goal. Knowledge of performance, or kinematic feedback, describes the nature of the movement pattern produced, such as limb position or timing, without necessarily stating whether the goal was met.

Telling a basketball player she missed a free throw is knowledge of results, while telling a diver that their tuck was not tight enough is knowledge of performance. Augmented feedback simultaneously drives four psychological and informational functions that influence learning and performance. Augmented feedback motivates the learner to increase effort, provides error information for corrections, directs attentional focus toward movement goals or internal execution, and creates a dependency that can cause performance to deteriorate when feedback is withdrawn.

Augmented feedback guides skill acquisition but must be managed carefully using reduced relative frequencies to prevent performance dependency.

Reducing the relative frequency of feedback during acquisition enhances long-term retention compared to providing feedback on every trial. While 100 percent feedback guides performance strongly during practice, it creates dependency and prevents learners from processing their own inherent feedback. Reducing relative frequency prevents this dependency and encourages independent error detection.

Winstein and Schmidt (1990) demonstrated in a limb patterning task that a group receiving feedback on 50 percent of trials learned more in retention tests than a group receiving 100 percent feedback across 400 total practice trials.

Frequency and Timing of Feedback

Continuous physical guidance and instantaneous feedback enhance practice performance but severely degrade retention and learning. Providing feedback instantaneously or restricting movements physically prevents learners from processing inherent feedback and developing error-detection capabilities, making them dependent on the artificial aid. Armstrong (1970) demonstrated that a physical guidance group performed best during practice on 3 days of lever movement, but worst on a no-feedback retention test, whereas terminal feedback groups showed superior learning. Postfeedback intervals of less than 5 seconds lead to planning issues.

Bandwidth and Summary Feedback

Bandwidth feedback withholds error information when performance falls within an acceptable tolerance zone and provides precise correction only outside that band. This method naturally produces faded feedback and increases the proportion of rewarding comments when performance improves, while stabilizing actions by eliminating minor trial-to-trial corrections. Sherwood (1988) found that a wider bandwidth of 10 percent produced more learning than narrower bandwidths of 5 or 1 percent using -1 to +1 mm and -2 to +2 mm tolerance bands.

You can apply bandwidth feedback rules during training:

• Do not give feedback if performance falls within an acceptable level or band of correctness.
• Tell the learner ahead of time that the absence of feedback means performance was essentially correct.
• Provide precise feedback indicating the amount and direction of error if performance falls outside the range of acceptability.

Summarizing a series of trials or providing only average feedback prevents dependency and enhances motor skill retention. Withholding feedback across multiple trials forces learners to analyze their own inherent feedback and filter out within-subject trial-to-trial variability, yielding more stable movement patterns. Schmidt, Lange, and Young (1990) found that a 5-trial summary feedback length was optimal for learning a batting-like task compared to 1-trial or larger summaries, while Young and Schmidt studied 10 swings observed before average feedback comments.

Conclusion

Mastering motor skills requires you to balance information processing limits, memory structures, and practice schedules. You must design environments that challenge cognitive processing without inducing overload. Permanent capability emerges only when you manage feedback schedules and practice variability to promote deep learning rather than temporary performance gains.

1. Structure practice schedules with high contextual interference to force active problem solving during skill acquisition.
2. Reduce the relative frequency of augmented feedback to prevent performance dependency and encourage internal error detection.
3. Distribute practice sessions across time to maximize consolidation and long term retention.
4. Scale task difficulty progressively to match learner capabilities and changing attentional demands.
5. Focus instructions on external environmental effects rather than internal body mechanics to leverage automatic movement control.

10 Key Ideas

1. Motor skills require achieving environmental goals with maximum certainty and minimum outlay of time and energy while relying on sequential stimulus identification, response selection, and movement programming stages.
2. Attentional resources are strictly limited and force a trade-off between controlled, effortful processing and automatic processing, while movement control integrates exteroceptive and proprioceptive sensory feedback loops.
3. Rapid actions are governed by generalized motor programs in an open-loop fashion, subject to physiological force variability, Fitts' Law speed-accuracy trade-offs, and temporal constraints.
4. Human motor performance depends on specific, independent, inherited abilities that shift in composition as learners transition from novices to experts through practice.
5. Distributing practice intervals across shorter sessions over more days and implementing variable practice schedules enhance long-term memory retention and schema development.
6. Random practice schedules and high contextual interference produce inferior performance during acquisition but vastly superior retention by forcing deeper cognitive processing and problem-solving.
7. Augmented feedback is supplemental information that drives motivation, provides error corrections, directs attentional focus, and creates performance dependencies.
8. Reducing the relative frequency of augmented feedback during acquisition enhances long-term retention compared to providing feedback on every trial by encouraging independent error detection.
9. Continuous physical guidance and instantaneous feedback enhance immediate practice performance but severely degrade retention and learning by preventing learners from processing inherent feedback.
10. Bandwidth feedback and summary feedback prevent dependency and enhance motor skill retention by withholding error information within a tolerance zone and forcing learners to analyze their own inherent feedback.