Practice¶

Element · All elements
Evidence · 32 claims (18 for, 10 mixed, 4 against) · 44 studies (19 causal, 11 quant-synthesis, 8 review, 5 theoretical, 1 qualitative), q1–q4 · 11 of 44 report an effect size · 19 claims rest on one study

Description¶

Practice asks learners to actively apply new knowledge or skills through structured activities rather than passively receiving information. It converts exposure into fluency and retention by requiring learners to generate responses, solve problems, or perform procedures under conditions that gradually approximate the target performance.

Design Implications¶

Practice is one of the most consistently supported instructional elements: active application produces better retention and transfer than re-reading or re-watching Retrieval Practice Improves Long Term Retention [+S]. Its effectiveness depends on structure — early practice should reduce unnecessary search for novices, for example by following worked examples Worked examples reduce unnecessary search for novices. [+M], and part-task practice can isolate difficult components before whole-task integration Part-task practice reduces cognitive load for novices. [+M]. Practice should be distributed over time and interleaved with feedback; massed, uncorrected repetition builds fluency in errors as readily as in correct performance.

Context¶

Requirements¶

Clear performance criteria so learners know what correct application looks like
Timely, specific feedback following each attempt (Feedback)
Sequencing that manages load: worked examples or part-task practice for novices, whole-task practice as competence grows (Fading)
Sufficient volume and spacing — one or two attempts rarely produce fluency

Constraints¶

Unguided problem solving early in learning imposes extraneous load and can underperform worked-example study Example-problem sequences reduce cognitive load compared to pure problem solving. [+M]
Practice without feedback entrenches errors; repetition alone does not correct misconceptions
Blocked, predictable practice produces strong short-term performance but weak retention and transfer compared to spaced, varied practice [~S]
Less effective when tasks exceed working memory capacity and no load-reduction supports are in place (Cognitive Load Management)

Target Learners¶

Novices in skill domains (STEM, vocational training, applied sciences) who need fluency before transfer Worked examples reduce unnecessary search for novices. [+M]
Learners building procedural fluency who benefit from part-task isolation of difficult steps Part-task practice reduces cognitive load for novices. [+M]
Advanced learners should shift toward whole-task, varied practice; excessive scaffolding at this stage reduces effectiveness Guidance becomes less effective as learner expertise increases. [~M]

Target Learning Goals¶

Procedural fluency: automating steps so working memory is freed for higher-order reasoning
Retention: strengthening memory through retrieval rather than review Retrieval practice improves long-term retention [+S]
Transfer: whole-task practice under varied conditions supports application to new problems Whole-task performance improves transfer. [+M]

Affordances¶

Deliberate Practice — practice enacts this principle when activities target specific weaknesses just beyond current ability, with immediate feedback and repetition with refinement rather than mere repetition
Active Learning — practice is the primary mechanism by which learners generate responses instead of receiving them, producing the generative processing passive formats lack
Retrieval Practice — practice from memory (rather than with open notes) converts each attempt into a retrieval event that strengthens long-term retention
Cognitive Load Management — sequencing practice from worked examples to faded to independent problems keeps load within working memory limits as expertise grows
Scaffolding — practice difficulty and support should fade in step with demonstrated competence

Claims¶

Design Decisions¶

Should practice ask learners to recall, or to review?¶

Default: make practice a retrieval attempt (recall, quiz, answer from memory) rather than rereading or rewatching. Testing beat restudy by g = 0.50 over 159 effect sizes, more at retention intervals of a day or more (g = 0.69) and with recall rather than recognition formats (0.61 against 0.29); a second meta-analysis found g = 0.51 across education levels, with multiple-choice as well as constructed-response items — Retrieval practice improves long-term retention [+S], Retrieval Practice Improves Long Term Retention [+S]
Default: include practice with feedback in computer-based lessons; undergraduates given lesson versions with practice scored significantly higher than those without — In one experiment with 256 undergraduates, computer-based lesson versions that included practice produced significantly higher posttest scores than versions without practice [+M]
Changes when: success is judged right after practice → restudying can look better; it was ahead at five minutes and behind at two days and one week, so do not evaluate a practice design on an immediate test — Retrieval practice improves long-term retention [~S]
Changes when: the goal is transfer to new questions → testing still helped (d = 0.40 over 192 effect sizes), most for application and inference questions and changed formats, but bias-corrected estimates often showed no transfer when none of the favourable moderators was present — Retrieval Practice Improves Transfer [~S]
Tested with: mostly university students on prose, word lists and facts in the laboratory; the second meta-analysis spans education levels; Pan & Rickard include medical-diagnosis problems.
Not settled: retrieval practice for procedural skills such as clinical procedures or multi-step algebra, where the claims' evidence is thin.

How hard should each practice attempt be?¶

Default: pitch items so most attempts succeed, or give feedback on every failed one. Without feedback, practice at 50% initial success or lower gave no testing effect (g = 0.03), and effects grew as success rose above 75% — Retrieval Failure Reduces Benefit [-M], A review reports that retrieval practice effects grow more robust as initial retrieval success rises, especially above 75%, and separately that giving more retrieval cues yields smaller effects [+M]
Changes when: feedback follows each attempt → lower success is tolerable; studies with feedback showed the largest effects regardless of initial success — Retrieval Failure Reduces Benefit [~M]
Changes when: retrieval is made too easy (more cues, or recall straight after study) → the benefit shrinks — A review reports that retrieval practice effects grow more robust as initial retrieval success rises, especially above 75%, and separately that giving more retrieval cues yields smaller effects [~M]
Tested with: a between-study moderator analysis of the testing-effect literature (Rowland 2014), reported again in one review chapter; not a direct manipulation of difficulty.
Not settled: a target success rate for adaptive systems; 75% is a moderator band, not a tested setting.

How should practice be scheduled?¶

Default: spread the same practice across sessions instead of massing it; spaced study raised recall from 36.7% to 47.3% over 271 comparisons, and a meta-analysis of 48 second-language experiments found a medium-to-large spacing effect — Spaced Practice Improves Retention [+S]
Default: set the gap by how long the learning must last; for 1,354 adults the best gap was about 20% of the test delay when the test was a few weeks away, falling to about 5% for a year (an upper-bound comparison with same-session review, d = 1.1) — Spaced Practice Improves Retention [+S]
Changes when: choosing expanding versus equal gaps → the shape matters less than spacing itself: expanding schedules showed no consistent advantage, and the L2 meta-analysis found them equivalent (one L2 vocabulary study found a small expanding advantage) — Expanding retrieval schedules have not shown consistent advantages over equally spaced or contracting schedules matched on total spacing [~M]
Changes when: the task is a motor or complex skill → the spacing benefit (d = 0.46 over 63 studies) varied with task type, and longer lags reduced effects for skills such as typing — Spaced Practice Improves Retention [~S]
Changes when: spreading L2 practice over years rather than months → it may be worse; this rests on one narrative review, and the meta-analysis did not test it — Distributed practice benefits L2 learning, and one review argues that spreading it over years can be worse than over months [~W]
Tested with: mostly young adults on verbal recall (714 of 839 comparisons in Cepeda et al. 2006), adults on an online panel, L2 learners (ages not established from the meta-analysis abstract), Japanese college students; supplemental spaced-repetition software nearly tripled delayed vocabulary credit for 22 EFL cadets (50.1% against 16.9%) — Supplemental computer-based spaced repetition activities nearly triple long-term vocabulary retention in EFL students compared with conventional instruction alone [+M]. Population-specific: evidence for children's long-term retention is thin.
Not settled: spacing for clinical procedures and school mathematics units; no claim here gives gaps for either.

Blocked or mixed practice?¶

Default: mix confusable problem types or categories in practice once each has been introduced; interleaving helped learners tell categories apart and pick the right strategy, and shuffled mathematics practice beat blocked practice on a delayed test despite worse performance during practice — Interleaving Improves Discrimination [+S]
Changes when: the material is word-based categories, or categories that are easy to tell apart → the benefit shrinks or reverses (overall g = 0.42; paintings 0.67, mathematics 0.34, word categories −0.39) — Interleaving category examples improves inductive category learning for visual and mathematical materials, but not for expository texts or word categories [~M]
Tested with: perceptual category learning (paintings, photographs) and mathematics practice; the claim pages do not record the learners' ages.
Not settled: how much blocked exposure should come first; the claims' Discussion sections advise it, but none of their evidence tests it. No claim tests interleaving of L2 grammar or clinical scenarios.

Examples first or problems first, and when to fade?¶

Default: for novices, start with a worked example before independent problems; example-first sequences beat problem-first ones on test performance with less mental effort, and alternating examples and problems was no better than examples alone — Sequencing worked examples with practice problems improves learning for novices [+M], Worked examples reduce unnecessary search for novices. [+M]
Changes when: transfer beyond the practised problem structure is the goal → algebra gains from examples were specific to problems of the same structure, so practice needs a wider range of problems — Studying worked examples improves novices' algebra performance on problems with the same structure as the examples [~M]
Changes when: learners gain expertise → withdraw the guidance; supports that help novices become redundant and can reverse for experienced learners, and fading worked-out steps gave better learning than problem solving alone or a block of examples followed by a block of problems — Worked-example guidance becomes less effective as learner expertise increases. [~M], Expertise Reversal Guidance Hurts Experts [~M], Fading support promotes the transfer of responsibility from instructor to learner. [+M]
Tested with: Dutch secondary students on circuit troubleshooting, Year 9, Year 11 and university students on algebra, electrical trainees reading circuit diagrams, a 71-student fading experiment.
Not settled: how to measure the point at which to fade; no claim here gives a threshold. No claim tests worked examples in L2 learning.

Part-task or whole-task practice?¶

Default: for absolute novices on a task with many interacting elements, practise isolated parts before the whole — Part-task practice reduces cognitive load for absolute novices during initial skill acquisition. [+M]
Changes when: transfer to professional practice is the goal → move to whole, varied tasks with less guidance; this is a design argument, not an experiment — A training-design argument, not tested by any study recorded here, holds that practising whole complex tasks improves transfer of complex skills [~W]
Tested with: one experiment on complex, high-element-interactivity material (Pollock et al. 2002); the whole-task side has no empirical entry.
Not settled: when to switch from parts to whole tasks.

How much practice, and practice until what?¶

Default: set a performance standard and let practice continue until it is met, rather than fixing the time; simulation with deliberate practice and mastery standards beat traditional clinical education (d = 0.71, 14 studies), and mastery learning programmes helped examination performance over 108 evaluations, more for weaker students — Simulation Based Education With Deliberate Practice Improves Clinical Outcomes [+M], Mastery Learning Improves Outcomes [+M]
Changes when: the course is self-paced → mastery programmes may add time on task and often lowered completion rates in self-paced college classes — Mastery Learning Improves Outcomes [~M]
Changes when: the domain lacks well-defined tasks and fast feedback → accumulated deliberate practice explained 18–26% of performance variance in games, music and sports but 4% in education and under 1% in professions — Accumulated deliberate practice is associated with performance, explaining a domain-dependent share of its variance: about a quarter in games, 4% in education and under 1% in professions [~M]
Changes when: judging dose by minutes per session → in one ASR pronunciation study practice frequency and session length showed no observable difference, but across ASR studies total programme length did (1–4 weeks g = 0.07, 5–8 weeks g = 1.01, 9 weeks or more g = 0.72) — Amount of ASR practice (days per week, session length) showed no observable difference in pronunciation improvement [~M]
Tested with: medical trainees on procedural skills (simulation), college, high-school and upper-elementary students (mastery), ESL/EFL learners using ASR, Korean English learners (dose).
Not settled: a minimum daily dose for any of these settings; no claim here gives one.

What should practice look like: drill or the target use?¶

Default: practise the skill in the form it will be used; grammar practice improved only the skill practised (comprehension or production), and task rehearsal improved fluency and complexity on the repeated task but not on a new task of the same type — Practice of grammar rules improves only the practiced skill, showing skill specificity of L2 automatization [+W], Task rehearsal improves fluency and complexity on the repeated task but does not transfer to a new task of the same type [~W]
Changes when: the practice is purely mechanical drill → argued to be of limited use, because it does not require learners to form a meaning and express it; automatization is argued to need extensive practice in communicative tasks in a consistent environment — Purely mechanical drills are of limited use because they do not train learners to conceptualize a meaning and express it linguistically [-W], Automatization of L2 skills requires extensive practice in a consistent environment, not repetition of grammar rules as in traditional exercises [+W]
Tested with: second-language learners, all reported second-hand through reviews and position papers (DeKeyser 1997 and Bygate 2001 are the primary studies behind them). Population-specific to L2; the same-structure limit on algebra examples above is the nearest evidence for mathematics.
Not settled: a controlled comparison of drill against meaningful practice; no claim here has one.

The evidence by question¶

After a model, does practice with support and feedback produce better unaided performance than no practice or unguided practice? In one experiment with 256 undergraduates, computer-based lesson versions that included practice produced significantly higher posttest scores than versions without practice [+M]
Does retrieving material (quizzes, practice tests, recall) improve later retention more than restudying it? Retrieval practice improves long-term retention [+S]
Does spreading study or practice over sessions improve later retention more than massing the same amount of it? Spaced Practice Improves Retention [+S]
Does training built around deliberate practice (mastery standards, feedback on each attempt, repetition) beat ordinary training? Simulation Based Education With Deliberate Practice Improves Clinical Outcomes [+M]

Demonstration — the typical precursor; observation without follow-on practice rarely transfers
Worked Examples — the low-load alternative to early unguided practice
Feedback — the mechanism that makes practice corrective rather than merely repetitive
Fading — governs the transition from supported to independent practice
Problem-Based Learning — a whole-task practice format for ill-structured domains
Hands-on Learning — physical manipulation as a practice mode in vocational and lab contexts

Patterns That Use This Element¶

Gagné's 9 Events — "eliciting performance" and "providing feedback" events
Merrill's First Principles — the Application principle: learners apply new knowledge with feedback and correction
Cognitive Apprenticeship — the practice phase following modeling and coaching

Examples¶

Use Worked Examples — Alternates fully solved problems with isomorphic problems for learners to solve; the practice problem immediately follows the demonstration.

Khan Academy — Practice exercise sets with immediate feedback, hint scaffolds, and spaced review built into the mastery system.

Duolingo — Distributed, interleaved practice with adaptive review scheduling; a large-scale implementation of spacing in practice design.

Codecademy — Learners write code in every lesson unit, practicing immediately after annotated demonstrations with automated feedback.

Key Sources¶

Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. doi:10.1111/j.1467-9280.2006.01693.x
Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363–406. doi:10.1037/0033-295x.100.3.363
Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4–58. doi:10.1177/1529100612453266
Sweller, J., & Cooper, G. A. (1985). The use of worked examples as a substitute for problem solving in learning algebra. Cognition and Instruction, 2(1), 59–89. doi:10.1207/s1532690xci0201_3

Page metadata

Field	Value
`type`	`element`
`id`	`practice`
`title`	`Practice`
`description`	`Encourages learners to apply new knowledge or skills through structured activities.`
`status`	`review`
`canonical`	`True`
`generated`	by `claude/unspecified` on `2026-08-29`
`trust`	unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.