Definition
Calibration is the alignment between how confident you feel that you know something and whether you can actually perform under retrieval.
Why it matters
Good calibration lets you spend study time where it matters; poor calibration lets confident illusion mask weak understanding.
Scientific basis
Studies of the 'illusion of knowing' show that students routinely overestimate their mastery, especially after rereading.
How to apply it
- Predict your test score before taking a practice test, then compare.
- Force yourself to retrieve before checking the answer.
- Treat the gap between confidence and performance as the most important signal in your study log.
Common mistakes
- Letting familiarity (re-reading fluency) inflate confidence.
- Avoiding self-tests to preserve a sense of mastery.
- Mistaking the absence of errors during review for evidence of mastery.
Examples
- A student predicts 90% on a self-test, scores 60%, and uses that gap to plan review.
- A pilot uses simulator scores to recalibrate her confidence in emergency procedures.
Related concepts
- Metacognition (complements)
Calibration is a specific metacognitive skill. - Illusions of Knowing (contrasts with)
Calibration is the antidote to illusions of knowing.
Where it shows up
Domains: student-learning, test-prep, professional-skills. Appears in the Metacognition & Self-Regulation cluster alongside 5 other concepts.
Source books
- Make It Stick: Brown, Roediger & McDaniel
Articles that use this concept (3)
- How to Tell How Much You Actually Know: how to know what you know
- Calibrate Your Confidence: metacognition calibration
- Design Self-Tests That Don't Lie: self testing design