Learners misjudge which learning strategies are effective¶

Claim · All claims
Evidence · 2 studies · 1 review r?, 1 associational r1 · q2
Currency · newest study 2013 · not yet checked against its sources · not reviewed by a person

Learners systematically overestimate the effectiveness of low-effort, fluency-based strategies (rereading, highlighting) and underestimate high-effort, generative strategies (retrieval practice, spacing, interleaving). This claim concerns metacognitive monitoring and strategy selection, not the effectiveness of the strategies themselves — those are covered in their own claim pages.

Subclaims¶

q2 i? College students overwhelmingly favor rereading and rarely choose retrieval practice as a study strategy, and most who do self-test say they do it to check what they know rather than because it produces better learning. → Karpicke Butler Roediger 2009

q2 i? A narrative review of the metacognition-and-studying literature concludes that most students believe rereading is a more effective learning strategy than self-testing, even though the reverse is true — a belief the review treats as a key driver of ineffective strategy choice. → Bjork Dunlosky Kornell 2013

q2 i? Surveyed on how they study, 177 students named rereading their most-used strategy; only a small minority reported self-testing, and those who did mostly described it as a way to check whether studying had worked rather than as a way to study. → Karpicke et al. 2009

q2 i? Learners misread fluency as learning: material that feels easy to process during study is judged better learned, which systematically favours massed rereading over spaced retrieval. → Bjork et al. 2013

q2 i? The misjudgement survives instruction and experience, so a system that merely offers retrieval alongside restudy should expect restudy to be chosen. → Bjork et al. 2013

Evidence¶

Karpicke Butler Roediger 2009¶

Karpicke, J. D., Butler, A. C., & Roediger, H. L. (2009). Metacognitive strategies in student learning: Do students practise retrieval when they study on their own? Memory, 17(4), 471–479. doi:10.1080/09658210802647009

q2 · survey study (open-ended and forced-choice questionnaires) · i? · no effect size reported (percentages of strategy endorsement, not a standardized effect) · n=177 college students (Question 1); 101 and 76 students respectively on the two versions of Question 2 · associational · r1

177 Washington University undergraduates were surveyed about their real-world study habits. On an open-ended free-report question, 83.6% listed rereading notes or a textbook as a strategy they use and 54.8% ranked it their #1 strategy, while only 10.7% (19 of 177) reported practising retrieval (self-testing) at all, and just 1.1% ranked it #1. On a forced-choice question asking whether they would reread, self-test, or do something else after reading a textbook chapter once, 57% chose to restudy and only 18% chose to self-test; of those who did choose self-testing, most explained they did it to check what they had learned rather than because testing itself produces better learning. The authors conclude students "lack metacognitive awareness of the testing effect."

Bjork Dunlosky Kornell 2013¶

Bjork, R. A., Dunlosky, J., & Kornell, N. (2013). Self-regulated learning: Beliefs, techniques, and illusions. Annual Review of Psychology, 64(1), 417–444. doi:10.1146/annurev-psych-113011-143823

q2 · narrative review of survey and laboratory studies · i? · no single effect size reported (synthesizes multiple studies' correlational and survey findings) · n=large (aggregated across the reviewed surveys and experiments, e.g. Kornell & Bjork 2007's 472 students and Karpicke et al. 2009's 177 students) · review · r?

This Annual Review of Psychology article synthesizes survey and experimental work on what students believe about how to study and how those beliefs relate to actual learning. It reports that when Kornell & Bjork (2007) asked 472 UCLA students why they used self-testing, only 18% said it was because they learn more from testing than rereading, versus about 70% who said they use it to check what they already know. The review states directly that "most students believe that rereading is a more effective strategy than is self-testing (McCabe 2011)," and frames this fluency-driven misjudgment, together with the dissociation between subjective ease and later retention, as a central reason self-regulated study choices go wrong.

Discussion¶

The fluency illusion. Learners judge learning by how easy and familiar material feels during study, not by how well it will be retrieved later. Rereading and highlighting make text feel fluent, producing high confidence and poor retention; retrieval practice feels effortful and error-prone, producing low confidence and strong retention. This dissociation between perceived and actual learning is the core reason learners misjudge strategy effectiveness.

Consequences for self-regulation. Because strategy selection typically rests on these faulty judgments, learners who believe they are studying effectively persist with ineffective methods. This undermines self-regulated learning: monitoring is inaccurate, so regulation decisions built on it are also inaccurate. Teaching learners to base judgments on retrieval performance rather than fluency is a promising corrective, but the evidence for specific calibration interventions still needs to be added to this page.

Boundary conditions. Misjudgment is strongest for novices, who lack the knowledge to evaluate their comprehension accurately; experts' judgments are better calibrated in their domain. Judgments made after a delay are more accurate than immediate judgments, which partly explains why cramming feels effective — the fluency is real at the moment of study but does not survive the delay to the test.

Design implications. If learners cannot be trusted to select strategies on their own, designers and instructors should default to effective structures — retrieval practice, spaced practice, and interleaving — rather than offering open study-strategy choice. Where choice is offered, it should be paired with strategy instruction that explicitly contrasts the felt ease of a strategy with its measured payoff, and with assessment for learning structures that give learners accurate feedback on what they actually know.

Open questions. The evidence recorded above shows what students prefer (rereading over self-testing) rather than measuring the judgment–performance gap directly; judgments-of-learning experiments that do measure it are recorded on the spacing and fluency claim pages.

Merged from “Learners Misjudge Effective Study Strategies” (learners-misjudge-effective-study-strategies): This claim concerns a metacognitive mismatch rather than a specific intervention: the strategies learners choose and the strategies that work diverge. The mechanism most commonly proposed is that learners use subjective fluency — how smoothly processing feels right now — as a proxy for learning. Strategies that make information immediately accessible (rereading, highlighting, massed practice) feel productive, while strategies that introduce desirable difficulties (retrieval practice, spacing, interleaving) feel ineffective precisely because they slow performance during study, even though they improve retention and transfer.

A key boundary condition is that the misjudgment is strongest for judgments made during or immediately after study, when fluency cues dominate. Judgments made after a delay, or based on actual test performance rather than felt ease, align better with objective learning. This suggests the misjudgment is not a fixed trait but a consequence of the cues learners attend to — a scope question that remains open in the literature: how much of the effect reflects poor strategy knowledge versus poor monitoring despite adequate knowledge.

Design implications. Because learners will not reliably self-select high-utility strategies, learning designs should build those strategies into the structure of instruction — scheduled retrieval, spaced review, interleaved practice — rather than leaving the choice to learners. Feedback that recalibrates judgments of learning is the primary corrective, connecting to Assessment for learning improves achievement. Where instruction reduces processing difficulty, designers should be alert to the flip side: reduced load can inflate fluency-based confidence without improving retention, so load reduction should be paired with retrieval demands rather than substituted for them — see Cognitive load reduction improves learning and Cognitive overload degrades learning.

Open questions. The relative weight of strategy knowledge versus monitoring failure is unresolved, as is the durability of calibration training: learners can be taught that effortful strategies work, but whether they maintain use of those strategies once instruction ends is not well established. Evidence entries supporting this claim still need to be added.

Merged from “Learners Misjudge Retrieval Benefit” (learners-misjudge-retrieval-benefit): The metacognitive mismatch. Learners tend to judge learning from the fluency of the current moment: rereading a passage feels smooth and productive, while attempting retrieval feels effortful and error-prone. Because perceived fluency is a poor proxy for durable learning, learners frequently select restudying over testing even when testing would produce substantially better delayed retention [-M]. This is a core instance of the broader illusion-of-knowing problem described under Fluency is a poor cue for actual learning. and connects to Learners' judgments of learning are often inaccurate..

Consequences for study behavior. When learners misjudge retrieval benefit, they allocate study time inefficiently — dropping self-testing early, massing restudy instead of spacing, and terminating study prematurely once material feels familiar [-M]. This undermines strategies that depend on learner-managed practice, such as retrieval practice improves long-term retention. and spaced practice outperforms massed practice.. The misjudgment also inverts the usual assumption behind learner-controlled environments: giving students more choice over study strategy can actively hurt retention when their strategy preferences are miscalibrated [-M].

Moderators and open questions. The misjudgment appears strongest before learners experience a successful delayed test; after receiving feedback on test performance, some learners update their beliefs toward testing [~W]. Whether brief experience with retrieval plus feedback is sufficient to durably correct beliefs, and whether the misjudgment generalizes across domains and learner ages, remain open questions pending evidence. The misjudgment is also likely strongest for novices, who lack the schema knowledge to distinguish transient fluency from durable learning — consistent with the broader pattern that strategy benefits and metacognitive accuracy both shift with expertise.

Design implication. Because learners cannot be relied on to choose retrieval voluntarily, designers should schedule retrieval and spacing into the learning environment rather than leaving the choice to metacognitive judgment [+M] — see cognitive-load-theory and self-regulated-learning for the tension between learner control and scaffolded scheduling. Where learner control is pedagogically required, pairing retrieval with immediate feedback and explicit explanation of the testing effect can partially correct the misjudgment [~W]. Low-stakes quizzing built into course structure (rather than offered as optional practice) sidesteps the misjudgment entirely while preserving retrieval's benefits.

Related Claims¶

• Self Assessment Accuracy Is Low Without Training — the general claim this finding bears on, with the studies behind it

• Learners underuse effective learning strategies such as spacing and retrieval practice — the general claim this finding bears on, with the studies behind it

• Retrieval practice improves long-term retention. — the generative strategy learners most often underestimate

• Spaced practice improves retention. — spacing feels harder than massing, so learners avoid it
• Interleaving improves discrimination and transfer. — blocked practice feels smoother but produces weaker learning
• Chunking reduces working memory load. — why fluent-feeling material can still be poorly retained
• Cognitive load theory — explains why effort and effectiveness dissociate during study
• Assessment for learning improves achievement. — external feedback can substitute for learners' faulty internal monitoring
• Self-regulated learning — the framework this claim constrains: inaccurate monitoring corrupts strategy regulation
• Cognitive load reduction improves learning — effective strategies manage load; learners' fluency judgments conflate low load with learning.
• Learner highlighting of text gives a small memory benefit but no reliable comprehension benefit, and evidence on other forms of annotation is not yet recorded here — annotating can support processing, but superficial marking is one of the strategies learners overprefer.
• Cognitive overload degrades learning — the flip side of the fluency trap: strategies that feel easy may not be the ones that manage load well.
• Fluent Illusions Mislead Self Assessment — related
• Instructions increased self-testing and cumulative learning but not study choices or feedback seeking in Experiment 1 — related
• Learners Misjudge Spacing Benefits — related
• Rereading is less effective than retrieval practice — related
• Rereading Is A Low Utility Study Strategy — related
• Most middle and high school students surveyed after classroom retrieval practice programs viewed them positively and said frequent retrieval practice helped them feel less nervous about exams — related
• In a teacher-run classroom survey, students most preferred practice test questions among three retrieval-practice strategies, and about three-quarters reported feeling well prepared after using retrieval strategies versus about a quarter without them — reports the opposite
• After training, d.tech teachers adopted low-stakes practice tests and spaced, interleaved practice in their classrooms — related
• Effortful, fluent-feeling experiences can create an illusion of learning: felt sense of learning is a poor gauge of actual learning — a narrower finding that bears on this claim
• Feedback interventions show phase-specific effects: advice feedback reduces reminder bias, but ranking feedback alters beliefs without increasing offloading — related
• Students rarely open transfer-practice cards immediately after solving; they want such problems organized into delayed, spaced wrong-book review instead — related
• Adult learners expect AI to enhance autonomy through choices, flexible learning spaces, metacognitive awareness, and structured self-assessment — related
• The interleaving literature is dominated by laboratory studies of university undergraduates, with no included study using high-school participants — related
• Predicting, using contextual analysis, and rereading were the most frequently used metacognitive behaviors; deciding when to stop, independent reading time, and testing were least frequent — related
• 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 — related
• genAI-supported work can improve while the constructive processing underlying durable expertise erodes (the performance–metacognition dilemma) — related
• Task completion with GenAI is a weak proxy for durable competence, with struggling learners at risk of an illusion of competence — related
• An early-semester presentation on fluency and feeling of learning improved students' attitudes toward active learning — a narrower finding that bears on this claim
• Explicit metacognition has severely limited access to underlying cognitive processes, and metacognitive reports can be inaccurate — related
• Training metacognitive self-regulation (self-testing plus selective restudy) produced greater paired-associate learning improvements in older adults than traditional encoding-strategy training when study was self-paced — related
• Students lack metacognitive awareness of the benefits of errorful learning: the three learning groups did not differ in perceived competence, judgments of learning, or rated method effectiveness even after tests — a narrower finding that bears on this claim
• Students rate their knowledge about cognition higher than their regulation of cognition — related
• Learners tend to feel restudying confers more learning than being tested even though the opposite is true, per the review — a narrower finding that bears on this claim
• No single metacognitive self-report can be recommended as the industry standard; the COMEGAM-ro has the strongest statistical support but is unreplicated — related
• AI-directed problem selection may diminish the metacognitive dimension of self-testing while preserving retrieval practice — related
• Learners overwhelmingly judge blocking more effective than interleaving even when their test performance favors interleaving — related
• Good learners do not always adhere to established practice principles; low contextual interference can still produce good learning — related
• Self-regulation limits supported a targeted teacher role: many students rarely thought before asking AI or could not tell when an AI explanation was wrong — a narrower finding that bears on this claim
• Undergraduate students reported using relatively effective strategies (elaboration, organisation, spacing) more than less effective ones, with elaboration most endorsed and self-testing least endorsed — related
• Reach — which lever settings remain in force when the student is alone with AI — is the framework's organizing question — related
• Strategies that reduce fluency and increase perceived difficulty (e.g., self-testing, spaced practice) often yield superior long-term learning, contrary to student beliefs — related
• Qualitative findings: students initially held inaccurate beliefs about strategy efficacy, and high effort from interleaved practice was misinterpreted as a sign of poor learning — a narrower finding that bears on this claim
• Individual performance feedback leads to more long-term choices for retrieval practice than general instruction on its benefits — related
• Students make substantial use of worked examples during self-regulated learning, but task selection patterns differ across studies and between high- and low-scoring students — related
• Generative activities improve students' monitoring accuracy but not their actual learning after restudy — related
• Ineffective learning strategies show substantially higher habit strength than effective strategies in a proof-of-concept survey — related
• There is currently no direct empirical connection established between teachers' learning about retrieval practice and their classroom implementation of it — related
• Students experience retrieval practice as more effortful than restudy, which leads to avoidance despite its long-term benefits — related
• Individual learners' self-awareness and metacognition of spaced practice is a distinct focus within spacing research — related