Self-Regulated Learning: Beliefs, Techniques, and Illusions

narrative review; the survey evidence it summarizes covers 472 (Kornell & Bjork 2007) and 324 (Hartwig & Dunlosky 2012) college students

Predominantly US COLLEGE students. Almost none of the metacognition survey evidence is K-12, which is the main limitation for a school-builder.

Annual Review of Psychology 64:417-444. A narrative review, not a computed synthesis — graded D on design and cited for the metacognitive-illusion evidence, not for any effect size. Its role in this database is to supply the citation for a claim two topics were already making: learners systematically misjudge which study methods work.

The illusion is measurable and large. Across Kornell (2009) experiments, 90% of participants performed BETTER after spaced than massed practice, yet 72% reported afterwards that MASSING had been more effective — subjective fluency during study is mistaken for learning. On retrieval: only about 18% of students who self-test say they do it because they learn more that way, while ~70% say they do it to gauge how well they have learned; most students believe rereading is the superior strategy (McCabe 2011). Only 11% spontaneously named retrieval practice as a study strategy in open-ended survey (Karpicke 2009), rising to ~42% when offered as a forced choice. 76% report rereading. 65-80% say nobody taught them how to study. The review's framing distinction — LEARNING versus PERFORMANCE — is why "desirable difficulties" feel bad while working.

Not applicable as an effect claim. Worth noting the reverse caution: the review's chapter on "overattributing differences in performance to innate differences in ability" is an argument about learner motivation, not a behavioural-genetic finding, and should not be read as evidence against heritability of ability.