JH
Jonathan Haber

Self-assess honestly — fail the card if you needed a hint

A card is only "known" when you produced a complete, correct answer with no hints.

Key takeaways

  • What it is: A card is only "known" when you produced a complete, correct answer with no hints.
  • Why it works: The effectiveness of any retrieval practice system depends entirely on accurate self-assessment. Rating a partial or hint-supported retrieval as "correct" removes that item from future review before it is actually consolidated, leaving a gap in long-term retention. Systematic self-inflation is the most common reason an apparently well-functioning retrieval practice system fails to produce the expected retention benefits.
  • Evidence: Backed by observational / correlational evidence.
  • Avoid: Retiring a card because you could recognize the answer when the back was shown — that is recognition, not retrieval, and it produces the illusion of knowing that breaks most self-study systems.

Why it works

The effectiveness of any retrieval practice system depends entirely on accurate self-assessment. Rating a partial or hint-supported retrieval as "correct" removes that item from future review before it is actually consolidated, leaving a gap in long-term retention. Systematic self-inflation is the most common reason an apparently well-functioning retrieval practice system fails to produce the expected retention benefits.

How to do it

  1. 1Before flipping a card or checking an answer, commit to a rating scale: only "pass" if you produced the complete correct answer without hints.
  2. 2Any partial answer, hesitation, or required hint = fail; keep it in rotation.
  3. 3If you rated something as "known" and then couldn’t retrieve it 3 days later, retroactively reset it to "in progress."
  4. 4Periodically sample "known" cards at random to verify your self-assessment is calibrated.

What the evidence says

Observational

Metacognitive accuracy — knowing what you know — is a well-studied component of effective learning. Overconfidence in recall is common and leads to premature termination of study. Calibration training improves both metacognition and subsequent learning outcomes.

Honest caveat: Self-assessment accuracy varies by individual and domain; there is no perfect calibration, only ongoing feedback. Periodic spot-checks of "known" material are the most reliable correction.

References
  • — Dunlosky & Rawson (2012), overconfidence produces underachievement: inaccurate self evaluations undermine students’ learning and retention, Learning and Instruction

Common mistake

Retiring a card because you could recognize the answer when the back was shown — that is recognition, not retrieval, and it produces the illusion of knowing that breaks most self-study systems.

IX Coach uses your response timing and self-rating together to assess confidence calibration — fast, confident responses that later fail in review are flagged and returned to rotation earlier than your self-assessment would have scheduled them.

Practice this with IX Coach →

Related practices