Metacognitive prompts improve learning¶

Claim · All claims
Evidence · 2 studies · 1 quant-synthesis r?, 1 review r2 · q3–q4 · i2 medium
Currency · newest study 2022 · 1 of 2 entries checked against their sources (automated) · not reviewed by a person

Prompts that direct learners' attention to planning, monitoring, and evaluating their own thinking can improve learning outcomes, particularly when embedded in structured learning tasks.

Subclaims¶

q4 i2 A meta-analysis of experimental studies in computer-based learning environments finds that metacognitive prompts (planning/monitoring/evaluation cues) produce a medium-sized improvement in learning outcomes relative to no-prompt control conditions, and a larger improvement in self-regulated-learning activity itself. → Guo 2022

q3 i? A systematic review of 35 studies of self-regulated-learning support in online learning environments judged prompting (14 studies) an effective way to enhance SRL strategies and learning performance, but reported no pooled effect, found that effectiveness varied with how prompts were implemented and with learners' prior knowledge, and did not examine publication bias. → Wong et al. 2019

Evidence¶

Guo 2022¶

Guo, L. (2022). Using metacognitive prompts to enhance self‐regulated learning and learning outcomes: A meta‐analysis of experimental studies in computer‐based learning environments. Journal of Computer Assisted Learning, 38(3), 811–832. doi:10.1111/jcal.12650

q4 · meta-analysis of experimental studies · i2 · medium effect, g=0.40, 95% CI [0.31, 0.49] · n=unreported in abstract (full text access-gated; k not stated) · quant-synthesis · r?

A random-effects meta-analysis of experimental studies conducted in computer-based learning environments (CBLEs) tested whether prompting learners to plan, monitor, and evaluate their own thinking during a task improves outcomes relative to unprompted control conditions. Metacognitive prompts significantly raised both self-regulated-learning activity (Hedges' g = 0.50, 95% CI [0.37, 0.63]) and learning outcomes (g = 0.40, 95% CI [0.31, 0.49]) compared to control. Moderator analyses found the effect varied with three features of the prompts themselves: whether they were paired with feedback, how task-specific they were, and whether they adapted to the individual learner — directly supporting this page's "prompt specificity" and "support fading/adaptability" moderator notes in the Discussion section below. The authors frame task-specific, individually adaptive prompting (with feedback) as the design implication for CBLEs.

Wong et al. 2019¶

Wong, J., Baars, M., Davis, D., Van Der Zee, T., Houben, G.-J., & Paas, F. (2019). Supporting Self-Regulated Learning in Online Learning Environments and MOOCs: A Systematic Review. International Journal of Human–Computer Interaction, 35(4–5), 356–373. doi:10.1080/10447318.2018.1543084

q3 · systematic review, narrative synthesis, no pooled estimate · i? · no pooled effect size reported · n=35 studies · review · r2

A systematic review of 35 studies of approaches to support self-regulated learning in online learning environments (23 of them at undergraduate level; searched April 2016), grouped by approach: 14 on prompts, 10 on integrated support systems, 2 on feedback, 4 on prompts combined with feedback. The prompting studies found more SRL activity (planning, goal specification, monitoring, evaluation) and better transfer, factual and problem-solving performance in several studies, but the authors say effectiveness "cannot be simply defined by one effect size" because prompts differed in form, intention, specificity and timing. In one study reviewed, lower-prior-knowledge learners benefited from prompts only once they had been trained to use them. Read in full (open access, Erasmus University repository).

"The evidence indicates that prompting is an effective way to enhance SRL and learning performance. However, the results should be interpreted with caution as publication bias was not examined."

Studies bearing on this claim¶

Do planning, monitoring and evaluation prompts help, and which ones? This table gathers the studies behind this claim and the 14 narrower claims that bear on it: 8 distinct studies (2 review, 2 associational, 2 causal, 1 quant-synthesis, 1 theoretical), q1–q4, 3 of 8 with an effect size coded. One row per study, codes as each entry records them; no effect is averaged and no verdict is drawn. Built by scripts/synthesize_claim.py from eval/synthesis/; a row that looks wrong is fixed on the claim it comes from.

Discussion¶

Scope and mechanism. Metacognitive prompts are typically short questions or cues (e.g., "What is your goal for this step?", "Does this answer make sense?") inserted into a task, worked example, or collaborative activity. They are expected to work by triggering self-regulatory processes that learners would not otherwise initiate spontaneously — see Self-regulated learning. The claim as stated is deliberately broad: prompts vary widely in content (planning vs. monitoring vs. evaluation), timing, and delivery. Two syntheses are recorded above: Guo (2022), a meta-analysis of experiments in computer-based learning environments (learning outcomes g = 0.40 against unprompted controls; its number of studies is not stated in the abstract read here), and Wong et al. (2019), a systematic review of 35 online-learning studies that pooled no effect. Both point in the claim's direction; neither covers classroom or paper-based prompting, and no primary study is recorded on this page yet.

Likely moderators. Several boundary conditions are plausible on general grounds. Guo (2022) reports that the effect varied with whether prompts came with feedback, how task-specific they were and whether they adapted to the learner, which bears on the specificity and fading points below; the others are untested here:

• Expertise. Novices are the most plausible beneficiaries, because they rarely engage in self-explanation or self-monitoring unprompted; for advanced learners, prompts may be redundant and consume working-memory capacity — consistent with the expertise-reversal pattern described in Cognitive load theory and Cognitive overload degrades learning.
• Prompt specificity. Generic prompts ("think about your learning") are less likely to help than task-specific prompts tied to the current step of the task. This parallels the distinction between generic and domain-specific guidance in Advance organizers improve learning.
• Timing and load. Prompts delivered during demanding problem-solving may interrupt productive processing rather than support it; prompts placed at natural pauses (before, between, or after tasks) are less likely to interfere with the load-management concerns captured in Chunking reduces working memory load.
• Support fading. Sustained benefits plausibly depend on gradually removing prompts as learners internalize the strategies; permanent prompting risks dependence rather than strategy acquisition — the same fading logic that governs Worked examples can become redundant or counterproductive for advanced learners.

Open questions. Whether effects persist after prompts are removed (i.e., whether learners internalize the strategies), whether effects transfer beyond the prompted task, and how prompt benefits interact with Activation of prior knowledge all remain to be established from the evidence base. A further open question is whether prompted reflection must be overt (written or spoken) to be effective, or whether covert cueing suffices — a distinction that matters for designs like 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 and 3-2-1 reflection.

Online environments (Wong et al. 2019). The online-learning review supports the direction of this claim but qualifies it: prompt effects depended on implementation and on learners' prior knowledge, and in the study it describes, lower-prior-knowledge learners needed training before prompts helped them. It pooled no effect, so it adds a second source for the direction, not for the size.

Related Claims¶

• 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 — annotation is a concrete strategy that externalizes the monitoring processes metacognitive prompts target
• Activation improves learning — prompts that activate prior knowledge are a closely related cueing mechanism
• Cognitive overload degrades learning — added prompts can themselves impose load if poorly timed
• Chunking reduces working memory load — relevant to when prompts fit within available working-memory capacity
• Self-regulated learning — the theoretical framework in which metacognitive prompting sits
• Reflective Practice Improves Outcomes When Structured — related
• Scaffolding improves learning — related
• Metacognitive strategy instruction, which includes comprehension monitoring, improves academic performance by a medium amount, and in one non-randomised study comprehension-monitoring training raised low achievers' reading scores — related
• Students showed deficiencies in maintaining and monitoring their reading plan within the three-element view of metacognition — related
• Metacognitive activity alone does not equate to effective regulation; unaccompanied metacognitive processes may be counterproductive — related
• The paper reports that metacognition—students thinking about how they understand their own learning—supports more efficient and effective learning and improved outcomes — a broader claim this one bears on
• AI feedback supports or undermines self-regulated practice depending on whether it supports or substitutes for internal monitoring — related
• Theory-guided five-step prompting of AI interaction improves open-ended explanation performance in construction education review, while unprompted AI access alone does not — a narrower finding that bears on this claim
• Scaffolding's effect did not vary by customization (fading/adding), change logic, or scope, and motivation scaffolds alone showed no significant effect — related
• Combining cognitive and metacognitive prompts optimizes germane processing; metacognitive prompts alone help only when remedial planning is prompted and remediation is possible — related
• Cognitive and metacognitive prompts in learning journals helped early in a term but became detrimental by its end; adaptive fading-out mitigated this — related
• Instructionally supported (prompted) journal writing outperforms unsupported journal writing on learning outcomes (Hedges's g = 0.78) — related
• Strategy prompts strongly increase learners' use of the prompted strategies, and the effect on comprehension is mediated by strategy use — 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
• Prior CBLEs for programming education have mostly supported SRL individually while disregarding the social context's influence on regulation — related
• Short-term prompting in weekly learning journals increases university students' self-reported metacognitive strategy use over time — a narrower finding that bears on this claim
• Rule-triggered pedagogical agents increased SRL activities and learning outcomes relative to untriggered agents — a narrower finding that bears on this claim
• Self-reported metacognitive regulation of reading contributes to expository digital text comprehension in a home e-learning environment — related
• Learning journal writing improves learning outcomes, but learners often do not apply sufficient strategies without support; prompts, worked examples and fading are the supports studied — related
• Teaching between study phases contributes to comprehension more than teaching after the study phase — related
• Higher compliance with MetaTutor's review-notes prompt is associated with larger proportional learning gains — a narrower finding that bears on this claim
• Higher compliance with MetaTutor's revise-summary prompt is associated with larger proportional learning gains — a narrower finding that bears on this claim
• Short-term prompting in weekly learning journals increases university students' self-reported cognitive strategy use over time — a narrower finding that bears on this claim
• Higher compliance with MetaTutor's suggest-subgoal prompt is associated with larger proportional learning gains — a narrower finding that bears on this claim
• Students' own cognitive and metacognitive activities predict learning, off-task activity lowers metacognitive knowledge, and group mates' relational activities raise domain knowledge — related
• Reflection increased in the adaptable and non-adaptable script conditions, but only the adaptable script had positive effects on monitoring, supporting overall positive effects on self-regulation — a narrower finding that bears on this claim
• More fixations on the learning content after an add-subgoal prompt are associated with higher learning gains, suggesting complying with that prompt may not be effective — a narrower finding that bears on this claim
• Compliance with MetaTutor's summarize, stay-on-subgoal, move-to-next-subgoal, and open-diagram prompts shows no significant effect on learning gains — related
• No behavioral measure of compliance with MetaTutor's stay-on-page and move-to-next-page prompts relates to learning gains — a narrower finding that bears on this claim
• In the generalized condition, compliance with several suggested actions was associated with higher essay scores, with one exception: note-taking after the last scaffold was associated with lower essay scores. — a narrower finding that bears on this claim
• The hypothesised mediation of prompt effects on self-efficacy via strategy use was not confirmed because prompts did not affect strategy use — reports the opposite
• Learners often fail to regulate their own learning in computer-based learning environments — related
• Self-regulated learning improves achievement — a broader claim this one bears on
• Strategy Instruction Improves Learning — a broader claim this one bears on
• Metacognitive Strategies Improve Learning — a broader claim this one bears on
• Cognitive and metacognitive prompts did not increase self-reported online strategy use, the quality of strategies in learners' notes, or learning outcomes in hypermedia learning — reports the opposite
• Students' use of metacognitive actions to direct group work increased over time when using the Knowledge Catalyser tools — a narrower finding that bears on this claim
• Personalized SRL scaffolds showed higher compliance than no and non-personalized scaffolds, linked to better essay scores — a narrower finding that bears on this claim
• Repeated reflective prompts may habituate, so scaffolding can replace rather than strengthen the user's own metacognitive habits — related
• Generative learning strategy prompts with metacognitive feedback yield higher self-regulation, strategy-use quality, recall and comprehension than strategy tools alone, controlling for prior knowledge — related
• AI introduced without pedagogical scaffolding can replace rather than reinforce learners' self-regulatory processes — related
• Pedagogical interventions and scaffolds increase learners' use of self-regulated reading strategies — related
• Fading-out prompts yield more cognitive strategy use and higher learning outcomes than permanent prompts — related
• Advanced learners may benefit from learner-controlled selection of learning tasks and levels of generality — related
• Cognitive-load-based quiz adaptation raised learning gains while domain-specific confidence-based adaptation did not — a narrower finding that bears on this claim
• Prompt compliance varied widely across the four metacognitive prompts, with interactive prompts responded to more appropriately — a narrower finding that bears on this claim
• Metacognitive prompting increases domain-general metacognitive SRL activities (goal specification and strategic planning) in young learners in a CBLE — a narrower finding that bears on this claim
• Prompted learners’ sequential SRL structure resembled traditional SRL models, using reflection more as an ending activity than unprompted learners — a narrower finding that bears on this claim