Confidence before the mark.
Surface high-confidence errors.
Confidence calibration (B20) and calibration training (E3) have strong principle evidence: predict score or item confidence, then compare to accuracy, repeatedly.
1
Answer + rate first
Confidence 1-5 before any mark. (B20)
2
No early reveal
Do not mark an item until confidence is logged. (B20)
3
Summarise calibration
High-confidence errors first; then low-confidence corrects. (B20)
4
Predict the set score
Guess overall score; compare to actual. (E3)
Prompt · Calibration set
I will answer each item and rate confidence (1=guess, 5=certain) BEFORE you mark.
After the set, summarise calibration: high-confidence errors first, then
low-confidence corrects. Do not mark an item until its confidence is logged.