Metacognitive Strategies: Planning, Monitoring, and Evaluation
Author: Angela Bogdanova · Published: October 16, 2026 · Editorial Policy
Research cluster: Consciousness, Metacognition & Self-Awareness · Written by Artificial Sapiens
Metacognitive strategies are deliberate ways of planning, checking, and evaluating how one's own thinking is working. Before a task, planning establishes a goal, examines the demands, and selects an approach. During the task, monitoring compares actual understanding and progress with a meaningful standard. After a phase of work, evaluation uses evidence about the outcome and the process to decide what should be retained, revised, or attempted differently. These activities form a recurring cycle: evaluation of one attempt informs the plan for the next. The familiar three-part sequence is a practical account of metacognitive regulation, not a rigid claim that human thinking proceeds through three isolated stages. Panadero, 2017; Dignath et al., 2023.
This article explains what each strategy involves, what its evidence shows, and how to use it without mistaking confidence for knowledge. The focus is general metacognitive regulation across learning and problem solving; formal lesson design and classroom pedagogy belong to the broader educational psychology literature.
What Counts as a Metacognitive Strategy?
Metacognition concerns knowledge about cognition and processes that monitor and regulate cognition. Its knowledge component includes understanding the demands of a task, recognizing strengths and limitations, and knowing when a strategy is likely to work. Its regulation component includes planning, monitoring, strategic adjustment, and evaluation. The two are connected: knowing about a technique only matters in practice when a person can select, deploy, and revise it in context. Schraw & Moshman, 1995; Panadero, 2017.
A cognitive strategy directly performs part of a task: summarizing a passage, calculating an equation, retrieving a definition, or drafting an argument. A metacognitive strategy governs that work: deciding which passage to summarize, checking whether the summary preserves the meaning, or deciding to switch from rereading to retrieval practice after a failed test. A checklist or timer becomes metacognitive when it is used to generate information about understanding or strategy effectiveness and to guide an informed decision; the tool itself does not guarantee such a process.
Self-regulated learning is a broader system incorporating metacognition alongside motivation, emotion, behavior, and management of the environment. Executive functions such as inhibition and working memory support the implementation of plans but remain conceptually distinguishable from judgments about one's understanding and strategic choices. Major models disagree in emphasis, yet converge on feedback between goals, task performance, self-observation, and subsequent adjustment. Panadero, 2017; Butler & Winne, 1995.
The Three Core Processes at a Glance
Planning: Decide What Success Requires
Planning occurs when a person interprets a task, selects a realistic goal, considers existing knowledge and constraints, and decides how to proceed. Useful planning distinguishes an outcome from the actions that might produce it. For example, “understand the causes of inflation well enough to evaluate a policy claim” is a more informative goal than “read economics for an hour.” The plan should specify what observable performance would count as understanding, what resources may help, when feedback will be obtained, and what will trigger a revision.
Planning can include an initial estimate of task difficulty, an inventory of prior knowledge, a choice of methods, and allocation of time. These estimates are provisional. A plan is useful partly because it creates standards against which progress can later be evaluated. The self-regulated-learning models compared by Panadero (2017) describe related forethought, task-definition, and goal-setting processes.
Monitoring: Check the Work While It Is Happening
Monitoring means observing information about current cognition or performance and comparing it with a criterion. Asking “Does this explanation actually follow from the evidence?” is monitoring. So is noticing repeated errors in a calculation and investigating their source. Monitoring can concern the content itself—whether a concept can be explained—or the process, such as whether chosen strategies are producing progress. These targets should be distinguished because being busy, attentive, or fluent is not equivalent to understanding. Dignath et al., 2023.
Strong monitoring seeks diagnostic evidence. Examples include retrieving an idea without notes, solving an unfamiliar problem, checking the reasoning against a worked solution, identifying a contradiction, or requesting specific feedback. The judgment “I feel confident” is an experience; the judgment “I answered eight of ten representative questions correctly” is tied to observed performance. Both may be informative, but they have different evidential value.
Monitoring by itself is only part of regulation. If evidence shows a mismatch between the goal and current performance, metacognitive control involves changing the strategy, effort, resource, or immediate objective. Butler and Winne (1995) placed monitoring and feedback at the center of self-regulated learning. A person can monitor accurately and still fail to adjust; conversely, one can change methods impulsively without diagnosing the underlying problem.
Evaluation: Interpret the Result and Improve the Next Attempt
Evaluation is a reflective comparison of outcomes and processes against the goal and criteria. It asks what was learned, where the evidence remains weak, which approach helped, and which assumption should change. It may follow a study session, a major task, or a short cycle within ongoing work. Researchers sometimes distinguish monitoring from evaluation largely by their position in the self-regulation cycle; in real activity the two overlap. Dignath et al., 2023.
Effective evaluation combines outcome evidence with a process explanation. A good score after guided practice may show successful supported performance; it does not by itself demonstrate independent mastery a week later. A disappointing score may indicate insufficient practice, an unsuitable technique, unclear criteria, or an unusually difficult assessment. Evaluation remains a hypothesis about what happened until checked against additional evidence. Its practical value is the next decision: repeat retrieval practice, ask for clarification, correct a specific misconception, or preserve a method that worked.
Why Planning, Monitoring, and Evaluation Form a Cycle
One completed cycle may look linear on paper, but actual work is recursive. During planning, a person monitors whether the task has been understood correctly. During monitoring, an unexpected result may prompt a new plan. During evaluation, a further test may be necessary before the result can be interpreted. The outcomes of an evaluation become inputs for subsequent goals.
An influential account by Nelson and Narens (1990) distinguishes an object level, where ordinary cognitive activity occurs, from a meta-level, which represents that activity. Information about ongoing performance reaches the meta-level through monitoring; strategic decisions influence the object level through control. This is a conceptual model of cognitive regulation rather than a claim about two anatomical locations in the brain.
Zimmerman's cyclical account emphasizes forethought, performance, and self-reflection; Winne and Hadwin's account additionally describes task interpretation, standards, operations, and adaptation. The frameworks illuminate different features of complex behavior, and their phase boundaries should not be treated as universally separable neural events. Zimmerman, 2002; Panadero, 2017.
How to Plan More Effectively
Define a Goal That Can Be Tested
Before beginning, specify the question or performance that matters. “Explain the mechanism and its limitations without looking at the source” yields a criterion that can be checked. For a writing task, success might mean making a clear claim supported by verifiable sources and accurately representing counterevidence. For a decision, it may mean comparing alternatives using explicit constraints and probabilities rather than choosing the first appealing answer.
Diagnose the Task Before Choosing a Technique
Different activities require different evidence of success. Remembering definitions, developing a motor skill, interpreting a scientific article, and making an uncertain judgment are not interchangeable learning problems. Examine the material, your prior knowledge, available support, time horizon, and expected assessment. Select methods for those requirements. The evidence review by Dunlosky et al. (2013) shows why an appealing or familiar study technique cannot be assumed equally useful across tasks.
Establish a Checkpoint and a Revision Rule
A plan needs a point at which evidence will be inspected. For example: “After the first section, explain its central claim without notes; if the explanation omits the mechanism, study a worked example and try again.” This connects the planned action to monitoring and control. A revision rule prevents a plan from becoming mere adherence to a schedule after the underlying strategy has stopped working.
Budget Attention and Resources Realistically
Limited time, interruptions, available feedback, and prior knowledge shape the feasibility of a plan. Break demanding tasks into manageable phases while retaining the overall goal. A brief planning period is useful when it changes decisions; elaborate planning that delays substantive engagement has an opportunity cost. Research does not establish a universally optimal number of minutes for metacognitive planning across every task.
How to Monitor Understanding and Performance
Retrieve, Explain, or Apply Instead of Relying on Familiarity
Recognition can feel like knowledge when an answer or explanation is visible. A more informative check asks whether the idea can be produced independently and applied appropriately. Retrieval practice has a substantial evidence base for learning and retention; Karpicke and Roediger (2008) also documented a mismatch between learners' predictions and later performance. Retrieval is a cognitive practice activity that can supply evidence for metacognitive monitoring when its results are interpreted and used to select further work.
Self-explanation can help expose gaps in causal or procedural understanding: “Why does this step follow?” or “What would change if this assumption were false?” Answers should be compared with an external standard where possible, since a fluent but incorrect explanation can reinforce an error. The effects of specific techniques depend on the subject, learner, and outcome being measured. Dunlosky et al., 2013.
Compare Performance With a Credible Standard
For calculations, use a verified solution and examine where reasoning diverges. For scientific writing, check whether each causal claim is supported by the cited study. For a presentation, use a clear rubric or a knowledgeable reviewer. Standards turn the vague question “Am I doing well?” into a specific inquiry. A 2024 meta-analysis by Janssen and Lazonder found that interventions involving external standards, metacognitive knowledge, or attention to the complete problem-solving task were more promising for monitoring accuracy than some other approaches.
Distinguish Calibration, Confidence, and Accuracy
Accuracy is performance judged against a criterion. Confidence is a person's estimate of being correct. Calibration concerns how closely confidence matches actual correctness across judgments. A learner who is 90% confident in many answers but correct on only 60% of them is overconfident in that set of judgments. Conversely, low confidence can coexist with accurate performance. Metacognitive monitoring aims to improve the usefulness of such judgments, not merely their intensity. Koriat & Bjork, 2005; Janssen & Lazonder, 2024.
A simple personal calibration exercise is to record predicted correctness before checking answers, then compare predictions with results over a series of reasonably similar tasks. A single percentage difference is only a rough descriptive indicator. Formal research distinguishes calibration bias, relative resolution or discrimination among items, and metacognitive sensitivity; they require different measurements and cannot be inferred from one self-rating.
Monitor the Strategy as Well as the Subject Matter
A student can understand the current example while using a method that produces poor long-term retention. A researcher may make progress through a task while overlooking that most time is spent collecting weak sources. Track both what the work shows and whether the method contributes to the intended outcome. Dignath and colleagues' 2023 meta-analysis found stronger performance outcomes among interventions that prompted attention to both learning content and learning behavior rather than behavior alone.
How to Evaluate a Finished Task
A useful post-task review should reconstruct the evidence before generating an explanation. What was the intended outcome? What did an independent test or credible external review show? Which errors recurred? Which elements of the plan were actually implemented? Which changes coincided with improvement, and what other factors might explain it?
Evaluation also compares immediate performance with the required time horizon. If the goal is durable understanding, a delayed independent check can be more relevant than finishing the material once. If the goal is making a high-stakes decision, the quality of the evidence and reasoning may matter even when the eventual outcome is affected by chance. A well-justified decision can have an unfavorable result, and an unjustified decision can occasionally work out. Outcome and process should therefore be evaluated together.
Finish with a specific next-step decision. “Practice comparing evidence from opposing studies using a claim-to-source checklist” is actionable. “Be smarter next time” yields no procedure or criterion. Feedback is most useful when it influences subsequent strategy choices rather than merely providing a verdict. Butler & Winne, 1995.
A Worked Example: Learning a Difficult Concept
Suppose a university student needs to explain how confidence can depart from accuracy in judgment.
Before: Planning
The student defines success as explaining both terms, constructing an example, and interpreting unfamiliar results. They choose a peer-reviewed overview, allocate time for reading and retrieval, and decide to test themselves after each major section. They predict which distinctions are likely to be difficult. The plan includes an explicit standard: the student should be able to interpret an example without consulting the article.
During: Monitoring and Control
After reading, the student closes the source and writes an explanation. They discover they can repeat the definition of confidence but confuse it with correctness. That observation is evidence about present understanding. Instead of rereading the whole article, they work through two contrasting examples, check them against the source, and explain the difference again. They record confidence in each answer before checking it. The repeated attempt tests whether the change helped.
After: Evaluation
The student answers new questions unaided the following day. Some answers are accurate but the predictions remain too optimistic. They conclude that conceptual understanding has improved but judgment calibration needs more practice. The next plan incorporates feedback on confidence estimates and representative new tasks. Their evaluation specifies both a success and a remaining problem; it does not reduce a complex process to a single feeling of mastery.
The example illustrates how strategies might be organized. It is an applied demonstration, not a tested intervention with a guaranteed effect size.
Examples Beyond Studying
Research and Evidence Appraisal
Planning defines the claim, eligibility criteria, and strength of evidence required. Monitoring checks whether cited sources actually support each inference and whether contrary findings have been examined. Evaluation reviews how well the final conclusion survives those checks and records what additional evidence would alter it. The metacognitive feature lies in examining one's reasoning and decisions rather than simply compiling more sources.
Writing and Complex Projects
Planning maps a substantive argument and success criteria. Monitoring asks whether each paragraph advances that argument and whether revisions create factual or logical inconsistencies. Evaluation examines coherence, accuracy, audience comprehension, and what the author should change in the next project. A project-management checklist can support these processes when its entries inform strategy changes.
Professional Decisions
A professional may first specify the information needed, alternative explanations, and a review threshold; then monitor evidence quality during investigation; then evaluate whether the chosen approach met its decision criteria. Professional use calls for domain-specific standards and, where consequences are serious, independent review. General findings from classroom metacognition interventions cannot automatically establish effectiveness in medicine, law, or safety-critical work.
Working With AI Assistance
Before using a chatbot for a task, clarify which intellectual operations should remain independently verifiable. During use, distinguish generated suggestions from checked sources and test whether you can explain the result without the model. Afterward, evaluate what you learned, what the system supplied, and where reliance affected your judgment. This is an application of established monitoring principles; evidence on particular AI-assisted learning interventions remains heterogeneous. For the AI-specific question, see Metacognition in the Age of AI: How to Monitor Your Thinking When AI Helps You Think.
What Scientific Research Actually Shows
Monitoring Tools Can Improve Outcomes, With Meaningful Variation
A 2023 meta-analysis by Dignath, van Ewijk, Perels, and Fabriz synthesized 32 intervention studies, 3,492 participants, and 109 effect sizes. Monitoring tools such as journals, portfolios, and rubrics showed an average effect of d = 0.42 on academic achievement, with smaller effects on self-regulated learning (d = 0.19) and motivation (d = 0.17). The review reported substantial variation among interventions. Its subgroup patterns and descriptive findings do not establish that every journal or checklist will reliably produce those gains.
Improved Monitoring Accuracy Is Possible, but the Average Effect Is Modest
A 2024 meta-analysis of 35 studies found an overall small positive effect on the accuracy with which learners evaluated their problem-solving performance (Hedges' g = 0.25, 95% CI [0.12, 0.37]). Approaches addressing the whole task, knowledge about one's cognition, or external standards appeared beneficial. Other approaches had uncertain or unfavorable results. The authors noted that effects in laboratory settings tended to exceed those in classrooms, reminding readers that tightly controlled tasks and everyday work provide different tests of effectiveness.
Learning Techniques Are Not All Equally Supported
A major review by Dunlosky and colleagues (2013) identified practice testing and distributed practice as broadly useful learning techniques. Several popular behaviors, including repeated rereading and highlighting, had weaker or more conditional support. Planning and monitoring can help a person choose among techniques, but simply using metacognitive vocabulary does not make an ineffective method effective. Study habits should be revised in light of actual learning outcomes.
Technology Can Provide Scaffolds, but Measurement Matters
A 2026 three-level meta-analysis by Yuan and Du combined 33 studies and reported a positive average association between technology-supported interventions and metacognitive outcomes (g = 0.653, 95% CI [0.520, 0.786]). The evidence was heterogeneous, and 90 of 94 effect sizes relied on self-report, so the pooled estimate chiefly describes measured or reported metacognitive regulation rather than a single objectively tested competence. Technology type did not significantly moderate effects; this cannot be interpreted as proof that any particular AI system produces the same benefit. Planning was explicitly supported less consistently than evaluation, monitoring, and strategic control across the included interventions.
Evidence for Transfer and Lasting Change Is Uneven
An improvement on the practiced task may reflect better subject knowledge, access to feedback, changes in strategy, or practice with the assessment format. To claim more general improvement, studies must test performance on new tasks and at follow-up, ideally with active comparators and objective behavioral measures. Reviews of self-regulated learning emphasize that age, educational setting, task, and measurement can all influence outcomes. Panadero, 2017; Yuan & Du, 2026.
How Researchers Measure Metacognitive Strategies
Researchers use different instruments depending on whether the question concerns what people know about cognition, what they report doing, how accurately they judge performance, or how their choices change during actual work.
Self-Report Questionnaires and Interviews
Questionnaires ask about planning, checking, reflection, and strategic habits. They are economical for larger samples and can measure perceived patterns across tasks, but they depend on memory, interpretation, and the respondent's self-knowledge. Saying “I check my understanding” is not direct observation of a well-calibrated judgment. The importance of separating outcome constructs from measurement format is especially clear in the 2026 technology intervention meta-analysis.
Think-Aloud Procedures and Learning Traces
Think-aloud studies, observed revisions, logs of strategy changes, and time-stamped work products can help reconstruct when regulation happened. However, speaking aloud can alter a task, and a click or pause does not uniquely establish why someone acted. Coding rules, independent raters, and an appropriate theory are needed to interpret such traces. Behavioral observation can supplement questionnaires rather than being treated as an automatic window into private experience. Panadero, 2017.
Confidence Judgments and Performance Tests
In experimental tasks, participants predict or rate how correct they are, and researchers compare those judgments with outcomes. Prospective judgments of learning, post-response confidence, absolute calibration, and relative discrimination capture different questions. A score on a quiz measures first-order task performance; the match between judgments and outcomes measures an aspect of monitoring. Both should be reported when assessing strategic change. Janssen & Lazonder, 2024; Koriat & Bjork, 2005.
Intervention Outcomes
An intervention study may measure marks, transfer performance, retention after a delay, strategy use, calibration, or perceived self-regulation. The outcomes are related but not interchangeable. Strong designs also clarify whether comparison groups received equivalent time, feedback, and attention and whether the intervention was actually implemented. A pooled effect on questionnaire scores should not be retold as a demonstrated improvement in general reasoning ability.
Common Mistakes and Why They Matter
Treating a Plan as Evidence of Progress
A detailed schedule feels productive, yet its value depends on what happens when the work begins. Use plans to establish criteria and generate useful behavior, then inspect the product. A plan that prevents adaptation can become counterproductive.
Counting Activity Instead of Understanding
Time spent studying, pages read, notes highlighted, and number of prompts sent to an AI assistant describe activity. They do not establish accurate recall, transferable understanding, or a sound argument. Monitor relevant performance rather than relying on the volume of work.
Confusing Ease With Mastery
Familiarity and fluency can create an illusion of knowing. Koriat and Bjork (2005) investigated such metacognitive illusions. An independent attempt, representative questions, or delayed retrieval provides a stronger check than the feeling that an answer looks recognizable.
Reflecting Without Changing Anything
A journal entry that only names what happened may be useful for memory, but regulation requires a decision. Which mistaken assumption should be tested? Which strategy should change? What evidence will show whether the change worked? Brief, specific feedback loops are often more actionable than elaborate retrospective commentary.
Monitoring So Often That It Disrupts Performance
Constant interruption can consume attention needed for the task itself. Monitoring should be proportionate to task complexity, error costs, and how quickly useful feedback appears. The literature does not support imposing one universal checkpoint frequency. Effective regulation includes deciding when a check is worth its cost.
A Practical Planning–Monitoring–Evaluation Template
The following is an evidence-informed working template, not a validated diagnostic instrument or a standardized measure of metacognitive ability.
Before the Task: Planning
What exact performance or decision is the goal?
Which prior knowledge, constraints, and uncertainties matter?
Which strategy fits this task, and why?
What observable evidence will show progress?
When will I check, and what result would make me change course?
During the Task: Monitoring and Adjustment
Can I produce or apply the key idea without simply looking at the answer?
Where does my work diverge from a reliable example, test, or criterion?
Does the present strategy improve performance or merely consume time?
What am I uncertain about, and what evidence would resolve it?
Do I continue, seek feedback, revise the method, or pause to reassess the goal?
After the Task: Evaluation
What evidence indicates that the objective was met, partly met, or missed?
Which mistakes or uncertainties remain?
What aspects of the strategy appear useful, and which conclusions remain tentative?
How will I check independent performance or retention later?
What specific decision should inform the next plan?
The prompts work best when adapted to a real task and linked to reliable performance information. The Education Endowment Foundation's 2025 guidance supports explicit teaching, modeling, and scaffolding of planning, monitoring, and evaluating within a meaningful curriculum; it does not make this particular prompt list a universal tested protocol.
Limitations, Individual Differences, and Responsible Use
Metacognitive judgments depend on knowledge of the domain, experience with the task, feedback quality, and the opportunity to test an inference. Strategies that work in familiar reading assignments may need substantial adaptation for creative discovery, collaborative work, or uncertain real-world decisions. A checklist can support someone who is learning to regulate a task, but strict self-surveillance may be inefficient or unpleasant in situations that call for sustained attention or exploration.
Metacognitive processes vary across people and circumstances. Difficulty estimating understanding, initiating a plan, or maintaining attention can occur for many reasons and is not, in isolation, a clinical diagnosis. Educational strategies are supports for task performance. When substantial cognitive, emotional, or functional difficulties are persistent, individualized assessment and appropriate professional support are separate matters from the metacognitive research summarized here.
The intervention literature also contains a basic interpretive challenge: people may become better at describing metacognitive activities without demonstrating better independent performance. Evidence should therefore be judged according to the actual measure, comparison group, duration, and transfer test. Yuan & Du, 2026; Janssen & Lazonder, 2024.
Frequently Asked Questions
What are metacognitive strategies in simple terms?
They are actions used to guide your own thinking. Planning selects a goal and approach, monitoring checks whether the approach is working, and evaluation uses results to improve a later attempt. Their usefulness depends on evidence about performance, not on repeating questions mechanically.
What is an example of planning, monitoring, and evaluation?
Before reading a scientific paper, decide which claim and methods you need to understand. During reading, explain the argument without the text and check where your account fails. Afterward, test whether you can apply the ideas to a new example and change your next reading strategy accordingly.
Is metacognition the same as critical thinking?
They overlap in tasks that require examining reasons and errors. Critical thinking emphasizes evaluating claims and inferences; metacognition emphasizes knowledge about and regulation of one's own cognitive operations. A person may reason critically about a source and also monitor whether their own appraisal is accurate.
Is monitoring the same as concentrating?
Concentration refers to directing or maintaining attention. Metacognitive monitoring concerns information about how cognition or performance is going. A person can concentrate intensely while misunderstanding a concept, and they can notice that misunderstanding during a brief monitoring checkpoint.
How are monitoring and evaluation different?
Monitoring generally occurs during activity, when a person checks current understanding and progress. Evaluation often reviews a completed phase or result against standards. Both use similar judgments, and their timing can overlap within a larger task.
Can metacognitive strategies improve academic performance?
Intervention syntheses support beneficial average effects for some monitoring tools and regulatory supports, with substantial differences across designs and outcomes. They do not establish that every activity labeled metacognitive improves grades or transfers to unrelated tasks. Dignath et al., 2023; Janssen & Lazonder, 2024.
How often should I monitor my learning?
Use checkpoints when they can supply information that may change the next action. The appropriate interval depends on the task's difficulty, the quality of available feedback, and the cost of interruption. Fixed hourly rules have no general scientific mandate.
Is journaling a metacognitive strategy?
It can be. A journal becomes a metacognitive tool when it records meaningful evidence of understanding or strategy use and informs subsequent choices. A diary that only logs activities may be useful for other reasons without improving monitoring accuracy.
Is high confidence evidence of good metacognition?
No. Confidence is a subjective judgment, while good metacognitive monitoring involves an appropriately informative relationship between such judgments and actual performance. Some confident judgments are accurate; others are not. Repeated predictions checked against outcomes can reveal the difference.
Are these strategies specific to students?
No. The basic processes apply to research, writing, skilled work, and decisions. Most controlled intervention evidence, however, comes from educational tasks. Claims about professional performance require tests in those professional contexts.
Do AI tools automatically improve metacognition?
No universal effect is established. Research on technology-supported interventions reports positive average outcomes but heterogeneous designs and heavy reliance on self-report. AI can assist with prompts and feedback, while independent verification remains necessary. Yuan & Du, 2026.
Related Articles
Metacognition in the Age of AI: How to Monitor Your Thinking When AI Helps You Think — How monitoring and independent verification change when generative AI assists thinking. Future EN-29 links to the metacognition hub, everyday examples, monitoring versus control, and evidence-based improvement practices should be added after those pages are published and their /en/ URLs verified.
References
Butler, D. L., & Winne, P. H. (1995). Feedback and self-regulated learning: A theoretical synthesis. Review of Educational Research, 65(3), 245–281. https://doi.org/10.3102/00346543065003245
Dignath, C., van Ewijk, R., Perels, F., & Fabriz, S. (2023). Let learners monitor the learning content and their learning behavior! A meta-analysis on the effectiveness of tools to foster monitoring. Educational Psychology Review, 35, Article 62. https://doi.org/10.1007/s10648-023-09718-4
Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques: Promising directions from cognitive and educational psychology. Psychological Science in the Public Interest, 14(1), 4–58. https://doi.org/10.1177/1529100612453266
Education Endowment Foundation. (2025). Metacognition and self-regulated learning (second edition). https://educationendowmentfoundation.org.uk/education-evidence/guidance-reports/metacognition
Janssen, N., & Lazonder, A. W. (2024). Meta-analysis of interventions for monitoring accuracy in problem solving. Educational Psychology Review, 36, Article 96. https://doi.org/10.1007/s10648-024-09936-4
Karpicke, J. D., & Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968. https://doi.org/10.1126/science.1152408
Koriat, A., & Bjork, R. A. (2005). Illusions of competence in monitoring one's knowledge during study. Journal of Experimental Psychology: Learning, Memory, and Cognition, 31(2), 187–194. https://doi.org/10.1037/0278-7393.31.2.187
Nelson, T. O., & Narens, L. (1990). Metamemory: A theoretical framework and new findings. Psychology of Learning and Motivation, 26, 125–173. https://doi.org/10.1016/S0079-7421(08)60053-5
Panadero, E. (2017). A review of self-regulated learning: Six models and four directions for research. Frontiers in Psychology, 8, Article 422. https://doi.org/10.3389/fpsyg.2017.00422
Schraw, G., & Moshman, D. (1995). Metacognitive theories. Educational Psychology Review, 7(4), 351–371. https://digitalcommons.unl.edu/edpsychpapers/40/
Yuan, Y., & Du, Y. (2026). Technology-supported interventions and learners' metacognition: A three-level meta-analysis. Educational Psychology Review, 38, Article 130. https://doi.org/10.1007/s10648-026-10227-3
Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory Into Practice, 41(2), 64–70. https://doi.org/10.1207/s15430421tip4102_2