Long-Term Memory¶

Learner Variable · All learner variables
Evidence · 116 claims (75 show a difference, 25 mixed, 16 no difference) · 73 studies (29 causal, 19 review, 13 quant-synthesis, 5 associational, 5 theoretical, 2 qualitative), q1–q4 · 26 of 73 report an effect size · 109 claims rest on one study

Description¶

Long-term memory holds information, such as words, facts, events, concepts and procedures, so that it can be retrieved and used later, over spans from hours to decades, and in principle indefinitely. Skills and knowledge get there through repeated practice and use, and stay accessible while they continue to be used. It matters for learning because what is stored there is the base for everything that follows: the prior knowledge and vocabulary that support reading and writing, and the arithmetic facts and procedures that free a learner to work on new mathematics problems. It develops through childhood and adolescence as the brain systems supporting it mature, and older learners lean more on memory-based strategies, for example in mathematics. It is measured with standardised memory batteries that ask learners to recall stories, lists or figures after a delay. It is distinct from working memory, which holds information briefly while it is being used; rehearsal and use move material from one to the other.

Implications¶

Context¶

Spaced, repeated practice and use is how skills and facts become durable.
Background knowledge and vocabulary carried in memory shape how well new reading and problems are understood.
Sleep, exercise and fewer distractions at retrieval all bear on what is stored and recalled.

Target Learners¶

Every age group; children building early reading and number facts, and adults drawing on accumulated knowledge.
Learners whose memory may be affected by trauma, sleep loss or poor health.

Target Learning Objectives¶

Retention of knowledge and skills over time.
Reading comprehension, writing, arithmetic fluency and problem solving that depend on stored knowledge.

Claims¶

A math teacher whose students had typically prepared for tests by rereading notes implemented three retrieval-practice strategies after learning that rereading is not effective for long-term retention [-W] — instruction changes it
Contextual diversity facilitates episodic recall but impairs recognition [~M] — instruction changes it
Declarative memory predicts immediate UGJT posttest gains for irregular past tense only under recasts [~W] — learners who differ on it differ in outcomes
Declarative memory shows no association with regular past tense learning under any CF type [-W] — learners who differ on it differ in outcomes
Deeper processing of target language improves lexical retention and long-term learning [+M] — instruction changes it
Desirable difficulty: harder-to-learn conditions yield enhanced long-term retention [+M] — instruction changes it
Digital game-based learning supports long-term retention of second-language vocabulary [+M] — instruction changes it
Emotion directly affects encoding, storage, and retrieval of information, including before awareness of the material, via broadening or narrowing cognitive resources and mood-dependent and mood-congruent processing [+W] — instruction changes it
Faster rate of learning may be negatively related to long-term retention (efficiency-effectiveness trade-off) [+M] — learners who differ on it differ in outcomes
Feedback on a multiple-choice test sharply reduces lure intrusions on a later cued recall test, with immediate and delayed schedules equally effective [+M] — instruction changes it
Having more multiple-choice alternatives benefits later retention when initial retrieval success is high but hurts learning when it is low [+M] — instruction changes it
In continuous recognition memory, the lag function is inverted U-shaped after a short retention interval and monotonically increasing after a long retention interval [~M] — instruction changes it
In paired-associate paradigms, the lag function is decreasing at very short retention intervals, inverted U-shaped at slightly longer intervals, and monotonically increasing at long intervals [~M] — instruction changes it
In the RANCH study, increasing road traffic noise exposure was associated with enhanced episodic memory, an unexplained counter-intuitive finding [+M] — instruction changes it
Inducing learners to recollect the prior study episode during retrieval practice enhances retention even when reexposure and initial success are held constant [+M] — instruction changes it
Initial retrieval conditions that provide less cue support, such as free recall rather than recognition or fewer letter cues, tend to produce better retention despite lower initial success [+M] — instruction changes it
Learners fail to recognize the benefits of pre-instruction testing, ranking it as least effective despite benefiting from it [+M] — instruction changes it
Meaningful background speech impairs delayed memory for texts in teenagers more consistently than transportation noise [+M] — instruction changes it
Passage cohesion showed no effect on delayed recall; means showed greater recall from less cohesive passages [-M] — instruction changes it
Passage familiarity (prior related knowledge) correlates .52 with recall, and only familiarity, not imageability, affected one-week delayed recall [+M] — learners who differ on it differ in outcomes
Patient history recall was near ceiling in both conditions, with significant decline over two weeks but no difference between genAI and ePBLM conditions [-W] — instruction changes it
PBL is superior to conventional instruction for long-term retention, according to a meta-synthesis of meta-analyses [+M] — instruction changes it
Practice tests benefit retention (g = 0.63) and transfer (g = 0.53) outcomes without a significant difference between them [+M] — instruction changes it
Practicing retrieval of some portions of an educational text can enhance retention of related nontested portions (retrieval-induced facilitation), but not when relational encoding is disrupted [+M] — instruction changes it
Procedural memory predicts immediate regular-past-tense UGJT posttest scores, controlling for pretest, only under metalinguistic prompts [+M] — learners who differ on it differ in outcomes
Procedural memory shows no association with irregular past tense learning under any CF type [-W] — learners who differ on it differ in outcomes
Recent research identifies time intervals during which spaced study should occur to maximize memory retention [+M] — instruction changes it
Repeated retrieval practice alters the rate at which learned items are forgotten, according to learn-to-criterion studies and forgetting-curve analyses the chapter reviews [+W] — instruction changes it
Repeatedly recalling a text without feedback or rereading produces the best one-week retention, while repeated studying leads on a final test at the end of the session [+W] — instruction changes it
Rereading is less effective than retrieval practice [+S] — instruction changes it
Retention trajectories diverged by centre: Centre 1 retained gains at 30 days while Centre 2 showed significant post-to-delayed decay [~W] — instruction changes it
Retrieval practice and spaced retrieval improve achievement and retention by medium amounts, while interleaving helps category learning only for some materials and blocking beats it for words [~M] — instruction changes it
Retrieval practice benefits extend to nonverbal visual materials, spatial map learning, and educational texts with different structures [+M] — instruction changes it
Retrieval practice benefits have been observed in children, healthy older adults, and memory-impaired patient groups, not only college students [+M] — instruction changes it
Retrieval practice benefits learners regardless of trait anxiety level, but higher trait or induced anxiety is associated with smaller testing effects [+M] — instruction changes it
Retrieval practice disrupted within-list order reconstruction of eight-item word lists, just as generation did [-M] — instruction changes it
Retrieval practice effects on mediator-cued final tests have been positive, but Coppens et al. (2016) concluded the true effect may be only about 0.10 to 0.20 [+M] — instruction changes it
Retrieval practice enhances final-test recollection of contextual details and speeds final retrieval, as shown by list discrimination, response time, and cumulative recall measures [+M] — instruction changes it
Retrieval practice improves long-term retention [+S] — instruction changes it
Retrieval practice instructions improved final recall of translations in Experiment 1 [+M] — instruction changes it
Spaced Practice Improves Retention [+S] — instruction changes it
Spaced Retrieval Improves Retention [+M] — instruction changes it
Spaced retrieval practice produces better final retention than massed retrieval even though spacing lowers initial retrieval success, and more absolute spacing enhances long-term retention [+W] — instruction changes it
Spacing out retrieval practice produces better long-term retention than massing it [+M] — instruction changes it
Supplemental computer-based spaced repetition activities nearly triple long-term vocabulary retention in EFL students compared with conventional instruction alone [+M] — instruction changes it
The argument that retrieval practice effects do not occur with materials high in element interactivity is contested, and the chapter judges its research base not convincing [~W] — instruction changes it
The GoldList Notebook Method increases long-term retention of L2 idioms compared with a vocabulary lesson alone [+M] — instruction changes it
The literature is supportive of recasts, but recast effectiveness varies with the language area and learner knowledge, focus, and aptitudes [~M] — an instructional effect differs with it
The pretesting effect on text memory grows with retention interval, being largest at one week [+W] — instruction changes it
The spacing benefit holds even at very short retention intervals, contradicting the Peterson paradox [+M] — instruction changes it
The timing of acute exercise relative to learning modulates memory consolidation [~M] — instruction changes it
VeriForge's integrated workflow is associated with higher one-week delayed recall of domain terms and a higher retention rate (exploratory) [+M] — instruction changes it
Vocabulary knowledge gained from eight encounters persisted over two weeks, with no significant immediate-to-delayed test differences [+W] — instruction changes it
Whether initial retrieval practice enhances semantic organization in final free recall is unclear, with results that may depend on how retrieval practice is implemented [~M] — instruction changes it
ANS acuity is not related to algebra achievement once covariates are controlled, whereas retrieval errors remain predictive [+M] — learners who differ on it differ in outcomes
Children with poor arithmetic fact abilities are significantly slower than matched controls only on the multiplication selection task [~M] — learners who differ on it differ in outcomes
Frequency of addition fact-retrieval errors predicts poor algebra equation memory but not coordinate plane or expression evaluation performance [~M] — learners who differ on it differ in outcomes
Less skilled comprehenders' inference failures stem more from incorrect premise recall, whereas skilled comprehenders fail more at integrating recalled information [~M] — learners who differ on it differ in outcomes
PoorAF children show a significant problem-size effect in multiplication RTs, with a marginally significant stronger problem-size influence than controls [~M] — learners who differ on it differ in outcomes
Semantic abilities indirectly predict word decoding efficiency in children with dyslexia via phonological awareness and rapid naming [+W] — learners who differ on it differ in outcomes
The comprehension-skill difference in inference making persists after controlling for memory for the text and delayed retention of the knowledge base [-M] — learners who differ on it differ in outcomes
Equating encoding time across presentation methods does not eliminate the forgetting-rate difference (Experiment 2) [+M] — instruction changes it
Equating total consolidation time across presentation methods equalizes forgetting rates (Experiment 3) [+M] — instruction changes it
Exercise performed four hours, but not immediately, after encoding enhances 48-hour retention of picture-location associations and hippocampal pattern similarity [+M] — instruction changes it
Lengthening consolidation time in the simultaneous condition reduces forgetting to a rate comparable to sequential presentation (Experiment 4) [+M] — instruction changes it
Sequential presentation of memory items produces less time-based forgetting than simultaneous presentation (Experiment 1) [+M] — instruction changes it
Short-duration anaerobic sprint exercise speeds vocabulary-type associative learning by 20% with increased serum BDNF that correlates with learning success [+M] — instruction changes it
Time-based forgetting occurs across presentation methods, with rate determined by consolidation time [+M] — instruction changes it
Sleep and consolidation support semantic memory development, suggesting delayed school start times may benefit memory [+M] — instruction changes it
An individual-then-collaborative activity structure (I+C) yields better delayed recall than collaborative-then-individual (C+I) and fully individual (I) structures [+W] — instruction changes it
Self-constructing a complete concept map produces better delayed recall than completing, sorting a map, or writing a summary, with the largest effect in the I+C structure [+M] — instruction changes it
Task type and activity structure interact on recall: self-constructed maps help mainly when individually prepared and then discussed in pairs [~M] — instruction changes it
No significant differences between conditions on the retention test, and scores of the Visual and Timeline conditions decreased significantly from post-test to retention [~W] — instruction changes it
Relationships between individual differences and final test performance were nonsignificant in retrieval practice conditions but significant in the rereading condition [~M] — learners who differ on it differ in outcomes
The self-study-before-discussion advantage persisted at one-week delayed tests, with lecture-condition scores declining over time while self-study-condition scores remained stable [+M] — instruction changes it
In Experiment 3, individual learners showed greater forgetting over 24 hours than paired learners [+W] — instruction changes it
Individual and paired flashcard learning yield comparable final cued-recall test performance [-W] — instruction changes it
Cued-recall testing can impair later memory for competitive, related but non-practiced information (retrieval-induced forgetting) [+M] — instruction changes it
Group testing shows advantages in some studies but mixed results overall for long-term retention and transfer [~M] — instruction changes it
Pretesting before instruction enhances learning of subsequently presented material even when pretest answers are wrong, with multiple-choice pretesting more effective than cued-recall pretesting [+M] — instruction changes it
Cue-only JOLs performed similarly to overt retrieval, and cue–target JOLs similarly to restudy, on Phase B memory and transfer (Experiment 1) [-M] — instruction changes it
Distributed practice outperforms massed practice for long-term learning, with a possible foreign-language exception [+M] — instruction changes it
Experts' complex movements are flexibly controlled via mental basic action concepts stored in long-term memory, as shown with tennis serves [+M] — learners who differ on it differ in outcomes
In Experiment 1, cue-only JOLs produced significantly better Phase B memory and transfer than restudy, while overt retrieval's memory advantage over restudy was only marginal [~W] — instruction changes it
In Experiment 1, interim tasks with high retrieval probability (overt retrieval, cue-only JOLs) enhanced new inductive learning more than low-retrieval tasks (cue–target JOLs, restudy) on both memory and transfer [+W] — instruction changes it
In Experiment 1, the interim learning task did not affect Phase A posttest performance, indicating the forward benefit was not due to improved prior learning [-M] — instruction changes it
Motor activities combined with high integration and high relevance produce the largest positive effects on memory, per a meta-analysis cited by the review [+M] — instruction changes it
Strategies that reduce fluency and increase perceived difficulty (e.g., self-testing, spaced practice) often yield superior long-term learning, contrary to student beliefs [+M] — instruction changes it
High and low CI practice do not differ significantly in overall relatively permanent gains (Δ pre-retention) [-M] — instruction changes it
Only low-working-memory-capacity students benefited from making pointing gestures when explaining number equations (review attributes this to Marstaller & Burianová, 2013) [~M] — an instructional effect differs with it
Both overt and covert retrieval practice outperform restudy when practice and test match (the testing effect) [+M] — instruction changes it
Non-interactive teaching does not enhance long-term retention relative to restudy at eight weeks [-M] — instruction changes it
Overt retrieval practice produces higher delayed recall of definitions than covert retrieval practice for learners studying declarative concepts [+M] — instruction changes it
Physics interest moderates long-term retention: low- and medium-interest students benefit from non-interactive teaching at eight weeks, high-interest students do not [~M] — instruction changes it
In spaced practice with post-feedback memory ratings, restudied Swahili–English pairs were recalled better than covertly or overtly retrieved pairs on a delayed final cued-recall test. [+M] — instruction changes it
Longer retention intervals between pre- and post-test weaken the effect of mental simulation [~M] — instruction changes it
Response format (overt vs. covert retrieval) did not affect final cued recall in spaced retrieval practice, and list structure (blocked vs. random; proportion of overt trials) did not moderate the effect. [-M] — instruction changes it
When the memory rating was removed entirely, retrieval practice was superior to restudying regardless of whether items were restudied or retrieved once, twice, or thrice. [+M] — instruction changes it
In applied settings, the beneficial effect of random practice on retention is almost negligible, as the competing meta-analysis itself reports [-W] — instruction changes it
Long-term nature exposures produced inconsistent effects on long-term knowledge acquisition, with positive garden and outdoor-lesson effects for some groups but not others [~M] — instruction changes it
The first phase of medical expertise development is accumulation of causal knowledge about disease and its consequences [+W] — instruction changes it
The third phase of expertise development uses episodic memories of actual patients in diagnosing new cases [+W] — learners who differ on it differ in outcomes
Two distinct implicit cognitive processes show different predictive links: implicit learning predicts skill learning, associative memory activation predicts language task success [+M] — learners who differ on it differ in outcomes
Studying with the material available during post-testing yielded higher post-test and delayed retention scores but worse transfer than retention-based studying [~M] — instruction changes it
Children preserve boundaries of separate learning episodes: hybrid statements combining content from two passages are rejected below chance [+M] — instruction changes it
Children retain self-derived knowledge over a one-week delay in classroom settings, with performance increasing from Session 1 to Session 2 [+M] — instruction changes it
Second-grade children self-derive new factual knowledge through integration of separate yet related learning episodes in classroom settings [+M] — instruction changes it
Self-derivation performance is positively correlated with stem-fact memory and predicts retention of self-derived knowledge one week later [+M] — learners who differ on it differ in outcomes
Self-derivation through memory integration extends to science curriculum content, with open-ended performance comparable to general-knowledge stimuli [+W] — instruction changes it
Surface-feature similarity of passage pairs does not affect self-derivation performance or patterns of episode-detail endorsement [-M] — instruction changes it
An authentic (greenhouse) video setting improves retention but not application of instructional video content in Experiment 1 [+M] — instruction changes it
Delayed-recall errors greatly exceeded immediate-recall errors, more so under the orderly procedure (4.33 times) than the random procedure (2.6 times) [+W] — instruction changes it
In delayed recall, the random procedure reversed the pattern, exceeding the orderly procedure by 16.4% fewer errors [~W] — instruction changes it
In Experiment 2 the authentic-setting effect on retention did not replicate and reinstating the setting as a test background produced no context effects [-M] — instruction changes it
In the reading fluency contrast model, numerical abilities and serial retrieval fluency predict multiplication fluency but working memory does not, indicating partial domain-specificity [+W] — learners who differ on it differ in outcomes
Numerical abilities and serial retrieval fluency predict third-grade multiplication fluency, while working memory shows no significant association in the baseline model [+W] — learners who differ on it differ in outcomes

Measures¶

Wechsler Memory Scale (WMS) (Wechsler, 2010): its logical memory subtest asks for details of a story heard once, as a measure of declarative memory.
Children's Memory Scale (CMS) (Cohen, 1997): assesses long- and short-term memory in verbal, visual and attention domains for ages 5 to 16.
Test of Memory and Learning (TOMAL) (Reynolds & Voress, 2007): measures verbal, nonverbal and composite memory for ages 5 to 59.

Adverse Experiences: childhood trauma is linked with changes in brain structures that form long-term memories.
Affect Regulation: emotion colours which experiences are stored and recalled.
Sleep: sleep consolidates newly learned material into long-term memory.
Working Memory: rehearsal moves information from short-term into long-term memory.
Executive Function: inhibition helps encode and retrieve the right information by suppressing competing meanings.
Prior Knowledge: relevant background knowledge helps new experiences get encoded.
English-Learner Status: bilingual learners may show an advantage in recalling verbal material.
Attention: attention guides what is retained, and distraction weakens retrieval.
Physical Well-Being: exercise before learning can help encoding.
Processing Speed: weak representations in memory slow retrieval and processing.
Academic and Critical Literacy: acquiring academic literacies depends on storing and retrieving what is learned.
Interest: curiosity can strengthen memory for new information.

Examples¶

Key Sources¶

Digital Promise. Learner Variability Navigator: Long-term Memory (Adult Learner, Building Literacy 4-6, Building Literacy 7-12, Building Literacy PK-3, Building Math 3-6, Building Math 7-10, Building Math PK-2, Portrait of a Learner 4-8, Portrait of a Learner 9-12, Portrait of a Learner PK-3). lvp.digitalpromiseglobal.org. Licensed CC BY-NC 4.0.

Page metadata

Field	Value
`type`	`learner-variable`
`id`	`long-term-memory`
`title`	`Long-Term Memory`
`description`	`The memory system that stores knowledge, skills and experiences over hours to decades so they can be retrieved and used later.`
`status`	`review`
`canonical`	`True`
`generated`	by `claude/unspecified` on `2026-10-16`
`trust`	unverified — no one has confirmed this page's content

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