Where it is used
Each entry identifies the volume, edition, chapter or appendix in which the work appears.
The evidence behind the series
Follow the research.
Explore every source.
Browse the scientific research and professional literature used across all seven volumes of The Science of Learning for Educators.
7book volumes
8author ranges
1searchable library
Search the collection
Find a paper, author or idea.
Choose an author range, search a title or concept, or narrow the collection by source type. Open any result to see where it appears in the books, read its evidence summary and follow available source links.
117 references
-
- Vol 5 RE: Chapters 8, App C
teacher feedback literacystudent feedback literacyfeedback process designevaluative judgementfeedback dialoguefeedback uptakefeedback workload managementJournal article Evidence summaryAt a glance
Carless and Winstone conceptualise teacher feedback literacy as the knowledge, expertise, and dispositions required to create feedback processes that students can use. Their framework contains three dimensions.
Key ideas
The design dimension sequences assessment and learning activities so students seek, generate, interpret, and apply feedback while developing evaluative judgement. The relational dimension treats feedback as communication requiring trust, sensitivity, dialogue, and attention to emotion. The pragmatic dimension concerns managing workload, disciplinary conventions, institutional constraints, and digital possibilities without abandoning learning priorities. Teacher and student feedback literacy are mutually dependent: teachers shape conditions for participation, while students' responses reveal where designs need refinement.
Why it matters
For educators, the article recommends moving effort from producing ever more comments toward designing opportunities for uptake, discussing purposes and responsibilities with learners, creating approachable relationships, and iteratively improving feedback practices from evidence of how students actually respond.
-
- Vol 4 RE: App C
retrieval practicetransfer of learningmemory representationsknowledge organisationknowledge elaborationdelayed applicationretrieval feedbackJournal article Evidence summaryAt a glance
Carpenter reviews evidence that retrieval practice can improve not only retention of previously tested information but also transfer to new situations. The studies considered vary the delay between learning and application, the format of practice and final tests, and the knowledge domain or context.
Key ideas
Across this emerging literature, retrieving information often produces advantages over additional study even when the later task is not identical to practice. Proposed explanations include strengthening accessible memory representations and encouraging organisation or elaboration that supports flexible use. The review is cautious because the transfer literature was still small and transfer can take many forms. For educators, the findings support asking learners to retrieve and apply ideas through varied questions, formats, examples, and delays rather than repeating one test type.
Why it matters
Retrieval should be designed around meaningful knowledge and followed by feedback when errors could otherwise persist.
-
- Vol 5 RE: Chapter 8
voluntary retrieval practiceself-directed studystudy strategy regulationretrieval promptsimplementation planningspaced retrievalmetacognitive monitoringstudy barriersJournal article Evidence summaryAt a glance
Carpenter reviews five kinds of intervention intended to increase students' voluntary use of retrieval practice during self-directed study.
Key ideas
The evidence shows that simply telling learners retrieval is effective, or even letting them experience a retrieval advantage, often does not change later choices. More promising approaches alter the decision environment: provide structured retrieval tools, prompt or require repeated use, give strategy training that includes implementation, support planning and monitoring, or reduce practical barriers to getting started. Students may otherwise use quizzes mainly to check knowledge, stop after one attempt, or prefer fluent restudy because it feels easier. The review emphasises that knowing a strategy and regulating its use are different achievements.
Why it matters
For educators, it recommends combining explanation with guided practice, convenient resources, spaced opportunities, feedback, and reflection on outcomes, then gradually transferring responsibility as productive study routines become established.
-
- Vol 2 RE: Chapter 3
spacing effectretrieval practicelong-term retentiondesirable difficultieslow-stakes recallretrieval feedbackstudy self-regulationdistributed practiceJournal article Evidence summaryAt a glance
Carpenter, Pan, and Butler review two well-supported learning strategies: spacing study over time and retrieving information from memory.
Key ideas
Both strengthen long-term learning across ages, content areas, and educational settings, yet learners often underuse them because massed study and rereading feel fluent and produce faster short-term gains. The authors examine mechanisms, boundary conditions, combinations with feedback and elaboration, and evidence from laboratories and classrooms. Spacing requires deciding how learning episodes are distributed; retrieval practice requires attempting recall rather than merely re-exposing oneself to material. Desirable difficulty can make effective study feel less successful while it is occurring, complicating self-regulation.
Why it matters
For educators, the review supports revisiting important ideas after delays, embedding frequent low-stakes recall, varying questions and contexts, correcting errors, and explaining why effortful practice is useful so students can make better independent study decisions.
-
- Vol 4 RE: Chapter 6
judgements of learningmetacognitive illusionssubjective fluencyactive learningdesirable difficultiesteaching evaluationsdurable learninglearning measurementJournal article Evidence summaryAt a glance
Carpenter, Witherby, and Tauber review illusions that distort students' judgements of learning and teaching effectiveness.
Key ideas
Smooth lectures, familiar material, engaging presentation, and low effort can increase subjective fluency and confidence without producing durable learning. Conversely, active learning, retrieval, spacing, and other effortful methods may feel less effective while creating stronger performance. These metacognitive errors affect study choices and can also influence course evaluations, which are further shaped by characteristics unrelated to learning. Heavy institutional reliance on satisfaction ratings may therefore reward practices that feel good immediately rather than those that develop knowledge and self-regulation.
Why it matters
For educators and leaders, the review recommends measuring learning independently of impressions, explaining desirable difficulties, giving students opportunities to compare predictions with later performance, and interpreting teaching evaluations cautiously alongside evidence of course design and achievement.
-
- Vol 2 RE: Chapters 11, 12
- Vol 3 RE: Chapter 13
- Vol 7 RE: Chapters 9, 10, 12, App D
universal design for learninglearner agencyengagementmultiple representationsaction and expressionlearner variabilityaccessible assessmentbelonging and inclusionGuideline Evidence summaryAt a glance
CAST's Universal Design for Learning Guidelines 3.0 provide design prompts for reducing barriers and supporting learner agency.
Key ideas
The framework organises options under engagement, representation, and action and expression, oriented toward learners who are purposeful and reflective, resourceful and authentic, and strategic and action-oriented. Version 3.0 strengthens attention to identity, belonging, bias, exclusion, interdependence, joy, and collective learning while retaining accessibility as foundational. The guidelines do not prescribe a checklist or separate accommodations for fixed learner types. They encourage anticipating predictable variability and offering purposeful options aligned with clear goals.
Why it matters
For educators and designers, the resource supports examining whose participation a goal, method, material, environment, or assessment may restrict; providing accessible ways to perceive information and demonstrate learning; supporting choice, collaboration, strategy development, and reflection; and revising designs from learner feedback rather than locating barriers within individuals.
-
- Vol 7 RE: Chapter 6
community of inquiryteaching presencesocial presencecognitive presenceonline discussioncritical thinkingdiscourse facilitationinquiry progressionJournal article Evidence summaryAt a glance
Castellanos-Reyes reviews two decades of the Community of Inquiry framework for online and blended learning.
Key ideas
The framework proposes that meaningful educational experience emerges through the interaction of teaching presence, social presence, and cognitive presence. Early research developed indicators and used content analysis to examine online discussion; later work produced and validated a survey instrument, expanded applications, and debated conceptual overlap, measurement, and the influence of course design. Teaching presence includes design, facilitation, and direction; social presence supports authentic participation and group cohesion; cognitive presence describes movement from triggering questions through exploration and integration toward resolution.
Why it matters
For educators, the retrospective recommends using the framework as a design and reflection tool rather than a formula: establish purposeful inquiry, create conditions for trust and communication, guide discourse toward integration, and examine whether activities actually support sustained critical thinking.
-
- Vol 5 RE: Chapter 15
lifelong learning scenarioslearning outcomes transparencylearning outcomes transferabilityqualifications recognitionlearning mobilityquality assurancescenario planninglearning pathway permeabilityReport Evidence summaryAt a glance
Cedefop presents five alternative scenarios for European lifelong learning to 2040, focusing on the transparency and transferability of learning outcomes.
Key ideas
The scenarios combine evidence about two decades of policy development with social, technological, economic, environmental, and political drivers. They are not predictions; they explore different balances between flexibility and permeability, on one hand, and stability and predictability, on the other. Each future raises consequences for learners, education and training providers, qualifications, recognition, mobility, quality assurance, and national and European coordination. The report uses scenario thinking to expose assumptions and trade-offs before decisions become fixed.
Why it matters
For policy makers and educational leaders, it supports stress-testing current arrangements, considering who benefits or is excluded under different pathways, improving recognition and accumulation of learning, and coordinating choices that allow people to enter, leave, and re-enter learning across institutions and life stages.
-
- Vol 1 RE: Chapters 2, 3, 5, 6, 7, 8, 11, 12, 13, App A, App B
- Vol 1 TE: App A
- Vol 2 RE: Chapter 10, App C, App D
- Vol 4 RE: App A, App C
- Vol 4 TE: App C
- Vol 5 RE: Chapters 7, App A
distributed practicespacing effectverbal recallretention intervalinterstudy intervallong-term retentionspacing optimisationexpanding schedulesJournal article Evidence summaryAt a glance
Cepeda and colleagues meta-analyse distributed practice in verbal recall, drawing 839 assessments from 317 experiments in 184 articles.
Key ideas
Spacing study episodes generally improves later retention compared with massing them together, but the best interval between study sessions is not fixed. Interstudy interval and final retention interval interact: as the time until the final test grows, the spacing that maximises performance also tends to grow. The review distinguishes the benefit of any spacing from the question of whether more spacing is always better and finds limited support for simple expanding schedules as a universal rule. Considerable variation and gaps in long-duration research remain.
Why it matters
For educators, the synthesis supports revisiting important material across separated occasions, choosing spacing in relation to how long knowledge should last, and avoiding both immediate repetition and excessively long gaps that make successful retrieval unlikely.
-
- Vol 1 RE: Chapter 8
spacing effectinterstudy intervalretention intervaloptimal retentiondelayed reviewretrieval schedulingknowledge accessibilityforgetting adjustmentJournal article Evidence summaryAt a glance
Cepeda and colleagues investigate how the interval between two learning sessions should relate to the desired duration of retention.
Key ideas
More than 1,350 participants learned facts, reviewed them after gaps ranging up to several months, and completed a final test after delays extending to one year. For each final-test delay, performance first improved and then declined as the interstudy gap increased, producing a ridge of optimal combinations rather than one universally best schedule. Longer desired retention called for a longer gap, although the optimal gap remained only a fraction of the test delay. The study demonstrates that recommendations such as “space practice” require a time horizon.
Why it matters
For educators and learners, the findings support planning reviews backward from when knowledge must remain accessible, avoiding immediate repetition, and using retrieval success and forgetting to adjust schedules rather than applying a single interval to every goal.
-
- Vol 7 PE: App D
cognitive load theorysplit-attention effectredundancy effectworking-memory resourcesintegrated instructional formatslearner knowledge effectsJournal article Evidence summaryAt a glance
Chandler and Sweller develop cognitive load theory through six experiments on instructional formats in electrical engineering and biology.
Key ideas
When learners must mentally integrate mutually referring sources, such as a diagram separated from the text needed to interpret it, working-memory resources are consumed by searching and coordination before learning can occur. Integrating essential information physically can reduce this split-attention burden and improve performance. Conversely, combining information that is independently understandable can create redundancy and add unnecessary processing. The effectiveness of a format therefore depends on the learner's knowledge and on whether each source is necessary, not on a universal preference for more explanation or multimedia.
Why it matters
For instructional designers, the article supports analysing the cognitive operations a learner must perform, integrating indispensable sources, removing or making optional redundant material, and testing designs with the intended learners rather than assuming that added information always helps.
-
- Vol 5 RE: Chapters 5, 6, 10, 11, App B
conceptual changebelief revisionmental-model transformationcategorical shiftontological categoriesmisconception diagnosismodel comparisonBook chapter Evidence summaryAt a glance
Chi distinguishes three forms of conceptual change according to the grain size and organisation of conflicting prior knowledge. Belief revision changes individual propositions that can be directly contradicted.
Key ideas
Mental-model transformation requires reorganising a network of connected beliefs and relations. Categorical shift is more radical: a concept has been assigned to an inappropriate ontological category, such as treating an emergent process as a material substance, so adding facts within the old category will not repair understanding. These forms differ in resistance and in the learning processes instruction must support. The chapter connects diagnosis with possible interventions, including refutation, model comparison, explanation, and helping learners recognise an alternative category.
Why it matters
For educators, the framework recommends identifying the structure of a misconception before responding, because a correction suited to an isolated belief may leave an underlying model or category unchanged.
-
- Vol 2 RE: Chapter 2, App D
- Vol 2 TE: Chi, M. T. H., & Wylie, R. (2014). The ICAP framework: Linking cognitive engagement to active learning outcomes. Educational Psychologist, 49(4), 219–243
- Vol 4 RE: Chapters 1, 2, 3
- Vol 5 TE: App C
active-constructive-interactive taxonomyactive engagementconstructive engagementinteractive engagementgenerative activitieslearner outputssubstantive dialogueJournal article Evidence summaryAt a glance
Chi proposes a learner-centred taxonomy for distinguishing active, constructive, and interactive activities by observable behaviour and hypothesised knowledge processes.
Key ideas
Active learners manipulate or select presented information; constructive learners generate an output that goes beyond it, such as an explanation, prediction, or concept map; interactive learners engage in substantive dialogue that builds on a partner's contribution. These categories generate the hypothesis that interactive learning is generally stronger than constructive, which is stronger than active, which exceeds passive reception, when activities are implemented as defined. The article reviews evidence, explains exceptions, and stresses that a task label is insufficient: notes, discussion, or technology can fall into different modes depending on what learners actually produce.
Why it matters
For educators, the framework supports designing prompts for generation and co-construction, examining learner outputs, and choosing comparison activities carefully when evaluating “active learning.”
-
- Vol 2 RE: Chapters 1, 2, 5, 7, 8, App A, App C, App D
- Vol 3 RE: Chapters 4, 5, 6, 7, App A
- Vol 4 RE: Chapters 1, 2, 3, 9, 20, 22, App B, App C
- Vol 4 TE: App C
- Vol 5 RE: App B
- Vol 7 PE: App D
- Vol 7 RE: Chapters 1, 2, 3, 4, 5, 8, 10, 11, App A, App B, App D
icap frameworkpassive engagementactive engagementconstructive engagementinteractive engagementknowledge-change processesco-construction of understandingJournal article Evidence summaryAt a glance
Chi and Wylie extend the earlier active-constructive-interactive taxonomy into the ICAP framework, adding passive engagement and specifying four observable modes.
Key ideas
Passive learners receive information; active learners manipulate it; constructive learners generate new inferences or representations; interactive learners co-construct understanding through dialogue that incorporates partners' ideas. The ICAP hypothesis predicts learning in the order interactive, constructive, active, and passive, subject to important constraints about task quality, partner contribution, prior knowledge, and implementation. The authors link each mode to possible knowledge-change processes and review laboratory and classroom evidence involving note-taking, concept mapping, and self-explanation. ICAP is a description of learner behaviour, not a label for a teaching method.
Why it matters
For educators, it provides a practical lens for redesigning tasks, checking what students actually do, prompting explanation or joint reasoning, and selecting valid comparison conditions when judging active-learning effects.
-
- Vol 1 RE: Chapters 2, 4, 5, 6, App B
- Vol 1 TE: App A
expertise in problem solvingdomain-specific knowledgeproblem representationdeep structural principlesqualitative analysisknowledge schemassurface-feature recognitionBook chapter Evidence summaryAt a glance
Chi, Glaser, and Rees synthesise research on expertise in problem solving, with particular attention to how knowledge is organised and used. Experts do not merely possess more facts or apply a general superior reasoning ability.
Key ideas
They recognise meaningful patterns, represent problems through deep domain principles, retrieve connected procedures efficiently, and use qualitative analysis to constrain a solution before calculating. Their advantages are largely domain-specific and arise from extensive structured knowledge and experience. Novices more often attend to surface features, search less selectively, and lack schemas that connect conditions with productive actions. Expertise can also bring limitations when familiar patterns are applied rigidly.
Why it matters
For educators, the chapter supports developing richly connected disciplinary knowledge, comparing surface-diverse problems with common structures, modelling problem representation and planning, and giving learners varied practice that helps them recognise when and why a principle applies.
-
- Vol 1 RE: Chapters 9, 12, App B
- Vol 2 RE: Chapters 8, 9
self-explanationgenerative inferenceprior-knowledge integrationmental-model revisioncoherent mental modelstransfer problem solvingexplanation qualityJournal article Evidence summaryAt a glance
Chi and colleagues test whether prompting students to explain a science text to themselves improves understanding.
Key ideas
Eighth-grade learners read a passage about the human circulatory system; those prompted to self-explain generated inferences, connected statements with prior knowledge, repaired gaps, and constructed more coherent mental models. They subsequently learned more and solved transfer problems better than students who read without the same explanation activity. Analyses showed that the quality and content of explanations mattered: productive self-explanation was not mere repetition, and learners differed in how extensively they revised their models. The work demonstrates a generative mechanism through which learners can make sense of incomplete instructional material.
Why it matters
For educators, it supports inserting prompts such as “explain why,” “how does this connect,” or “what follows,” modelling useful explanations, and checking their substance rather than counting how often students speak or write.
-
- Vol 1 RE: Chapters 6, 11, App A
physics problem categorizationproblem representationexpert-novice differencessurface featuresunderlying principlessolution structuresdisciplinary knowledge organizationJournal article Evidence summaryAt a glance
Chi, Feltovich, and Glaser compare how physics experts and novices categorise and represent mechanics problems.
Key ideas
In sorting and think-aloud studies, experts grouped problems by underlying principles and solution structures, such as conservation of energy or Newton's laws. Novices relied more heavily on literal surface features, such as whether a problem mentioned an inclined plane, spring, or pulley. The difference was not simply the amount of knowledge: experts' categories activated connected conceptual and procedural knowledge that guided solution planning, while novice representations often lacked this organisation. Surface cues can still affect experts, and developing expertise is gradual.
Why it matters
For educators, the findings support teaching students to identify governing principles before selecting equations, comparing problems with different contexts but common structures, explaining categorisation decisions, and using sorting tasks to reveal how learners organise disciplinary knowledge.
-
- Vol 6 PE: App E
instructional roundsinstructional coreproblem of practicenon-judgemental observationevidence-based debriefingprofessional learning networkscollective instructional capacityBook Evidence summaryAt a glance
City and colleagues adapt medical rounds into a disciplined network process for learning from classroom instruction.
Key ideas
Participants identify a focused problem of practice, observe multiple classrooms without evaluating individual teachers, describe evidence in specific non-judgemental language, identify patterns, predict what students are learning, and propose a next level of work connected to the organisation's improvement strategy. The instructional core—the relationship among student, teacher, and content—keeps attention on tasks and learning rather than personalities or generic impressions. Repeated rounds help educators develop a shared language and collective capacity to see instruction more accurately. The process is not a walkthrough, compliance inspection, or one-time event.
Why it matters
For leaders and teacher networks, the book recommends clear norms, confidentiality, careful observation training, evidence-based debriefing, and sustained follow-through so that rounds inform coherent professional learning and system action.
View or buy the cited work >>Read a similar article free online >>
-
- Vol 7 RE: Chapter 3
extended mindactive externalismparity principlecognitive systemsexternal memorycoupled cognitive systemsdistributed cognitionJournal article Evidence summaryAt a glance
Clark and Chalmers argue for active externalism: under appropriate conditions, parts of the environment can constitute a cognitive process rather than merely supply inputs to an internal mind.
Key ideas
Their parity principle asks whether an external process would be accepted as cognitive if it occurred inside the head. The contrast between Inga's biological memory and Otto's reliably consulted notebook illustrates how information that is constantly available, automatically endorsed, and easily accessible can function as memory. Language, diagrams, tools, and collaborative arrangements can form coupled systems that accomplish thinking. The claim is stronger than saying tools influence cognition; it concerns where a cognitive system is bounded.
Why it matters
For educators, the paper invites attention to how learners coordinate with notes, representations, devices, and people, while also raising questions about reliability, access, agency, and what happens when an external resource is removed.
-
- Vol 2 RE: Chapter 14
- Vol 7 PE: App D
- Vol 7 RE: Chapters 1, 10, App B, App D
media comparisoninstructional methoddelivery mediumcausal attributiontreatment effectsmethodological confoundsinstructional efficiencyJournal article Evidence summaryAt a glance
Clark reviews media-comparison research and argues that no delivery medium by itself produces learning.
Key ideas
Apparent advantages for television, computers, or other formats can often be explained by differences in instructional method, content, effort, time, or novelty that were confounded with the medium. A medium can affect cost, access, convenience, and the methods that are practical, but causal claims about achievement require holding the instructional method constant. The article also critiques theories that attribute learning directly to media symbol systems without isolating the psychologically active treatment. Its provocative vehicle analogy stimulated a long-running debate, and the conclusion should not be read as saying design or technology choices are irrelevant.
Why it matters
For researchers and educators, the paper recommends identifying the instructional mechanism, comparing equivalent methods, measuring efficiency as well as outcomes, and avoiding claims that a device or platform improves learning merely because it carries an effective activity.
A transparent research trail
More than a bibliography.
What it contributes
Evidence summaries explain the central idea and why the source matters to educators.
How to access it
Entries link to lawful full text where available, or to a publisher, library or purchase page.
Edition abbreviations
- TE
- Teacher Edition
- PE
- Practice Edition
- RE
- Research Edition
Citation checks
References are checked against publisher records, scholarly indexes, DOI or ISBN metadata, repositories and author records where available. An entry marked [Citation not verified] preserves the wording found in the relevant volume without attributing findings to an unconfirmed work.
From sources to a coherent pathway