Metacognition, thinking about one's own thinking, is the skill that decides whether the other strategies are actually used. Well-calibrated learners know what they know and do not know, and adjust their study accordingly. Two systematic biases distort this calibration. The Dunning-Kruger effect causes beginners to overrate their competence because they lack the expertise needed to recognize gaps. Overconfidence bias, more generally, confuses familiarity with mastery, the same illusion that drives re-reading and highlighting. Calibration improves most reliably through frequent, low-stakes self-testing, which provides concrete feedback on what is actually known versus what merely feels known. The illusion of knowing is particularly insidious because repeated passive exposure to material, the mere exposure effect, increases familiarity and the subjective feeling of mastery without improving actual recall. Passive strategies such as highlighting and re-reading produce almost no retention gain relative to the effort invested, while creating a false sense of progress.
The Feynman Technique operationalizes metacognition into a four-step workflow. First, choose the concept and study it. Second, explain it in simple language as if teaching a twelve-year-old. Third, identify any gaps in the explanation and revisit the source material to fill them. Fourth, simplify the language further and review the result. The technique works because it forces active retrieval, exposes knowledge gaps, promotes deep processing, and demands simplification, all of which together surface misunderstandings that passive reading hides. A practical implementation is the "blank paper" test: write the topic at the top of a blank page, then explain everything you know without referring to notes; any gaps the test reveals indicate areas needing more study. In mathematics, the technique can be applied by narrating each step aloud in plain language ("I'm dividing both sides because...") and stopping wherever a step cannot be explained simply. The technique has limits: it works less well for purely rote material such as lists and dates, and it can oversimplify nuanced topics, so it pairs best with techniques suited for factual recall, like spaced repetition.
Self-explanation and elaborative interrogation are closely related and often confused. Self-explanation is the broader act of generating inferences about how new information works, why it is true, and how it connects to prior knowledge; Michelene Chi and colleagues first documented the self-explanation effect in 1989 in studies of physics students. Summarization is different: summarization restates information in fewer words, while self-explanation involves inference, generating new connections and filling in reasoning steps rather than paraphrasing. Elaborative interrogation, formalized by Keith Pressley and colleagues building on John Bransford's earlier work, is often considered a focused subtype of self-explanation that specifically targets causal "why" reasoning. By forcing causal connections between new facts and prior knowledge, elaborative interrogation builds flexible mental models that can be applied to novel problems rather than memorized facts tied to specific contexts. When applied during reading, both techniques are simple: pause after each paragraph or example, ask what the main idea is, why each step follows, and how it connects to what you already know, then write or say the answer before continuing. Combining self-explanation with retrieval practice and interleaved practice produces additive benefits, since each technique targets different cognitive processes.
Teaching is one of the most reliable ways to learn. The protégé effect, a term popularized by John H. Falk and Lynn D. Dierking, refers to the tendency to encode material more deeply when one expects to teach it, even if no teaching ever takes place. The cognitive mechanisms are the same as in self-explanation: preparing to teach triggers deeper processing, retrieval practice, organization of knowledge, and anticipation of questions. In a study group, when each member prepares to teach a portion of the material, everyone benefits from deeper encoding, anticipated peer questions, and retrieval practice. Importantly, members who feel less expert often gain the most from having to explain, because doing so surfaces and repairs their weakest representations. Crucially, all four techniques depend on active retrieval: pulling information out of memory rather than re-exposing yourself to it, which strengthens memory, identifies gaps, and improves long-term retention more than passive review.