Recall before recognition

Test yourself by trying to produce an answer freely before checking multiple-choice options.

Why it works

Recognition is easier than recall because it requires only matching an incoming stimulus to a stored trace — the answer is present in the environment. Recall requires generating the answer from internal memory alone. FOK judgments tend to predict recognition success well because they are partly based on availability of partial traces that recognition can match. Testing by recall first prevents the recognition option from inflating your confidence about what you can actually produce.

How to do it

1. When reviewing material, cover the answer and try to write or say it before looking.
2. Only after a genuine attempt (30+ seconds of effort) consult the answer or a multiple-choice format.
3. Score yourself on recall, not recognition — a correct recognition after failed recall is a gap, not a hit.
4. Return material you recognized-but-couldn’t-recall to active study; do not move it to the "known" pile.

Evidence

Metcalfe and colleagues documented that FOK judgments correlate more strongly with later recognition than with later recall — demonstrating that confidence can be genuinely predictive for one memory measure while misleading for another. (observational)

FOK accuracy varies considerably across individuals, domains, and material types; the systematic gap between FOK and recall performance is robust but not universal.

Sources

• Metcalfe & Shimamura (1994), Metacognition: Knowing About Knowing (MIT Press)
• Metcalfe (1993), "Novelty monitoring, metacognition, and control," Journal of Experimental Psychology

Common mistake

Studying by recognition (multiple-choice, flashcard flip-to-reveal) and feeling confident because you recognize answers — which reflects a high FOK but not the recall ability you will need under test conditions.