Changes in students’ confidence calibration across a sequence of low-stakes confidence assessments

In this paper, we explore changes in secondary school mathematics students’ (age 12–13, N = 203) confidence calibration (i.e., the correlation between accuracy and confidence) across a sequence of six low-stakes (i.e., not-for-credit) formative assessments administered over 2 months during normal classroom teaching. For each item in each of the assessments, students were asked to state on a 0–5 scale how sure they were that the answer that they had given was correct. Afterwards, they calculated their total score in a manner intended to incentivize improved calibration: the sum of the confidence ratings on the items that they answered correctly, minus the sum of the confidence ratings on the items that they answered incorrectly. Students broadly improved their calibration across the assessments, and a qualitative analysis of their views on the practice of confidence assessment was overwhelmingly positive.