Summative self-assessment accuracy in higher education: An exploratory study of a grade incentive, individual differences, and perceived versus actual accuracy

This exploratory quasi-experimental study examined whether an accuracy-based incentive for summative self-assessment improved alignment with instructor-assigned grades in Spanish higher education, which student characteristics were associated with accuracy, and whether perceived accuracy corresponded to actual accuracy. A total of 142 second‐year undergraduates from seven classroom groups were assigned to an incentive or control condition, completing rubric-based self-assessments and questionnaires on self-regulation, self-efficacy, and prior academic achievement. No clear evidence of improved accuracy was found; Bayesian and equivalence analyses indicated the data were inconclusive rather than a null effect. Prior achievement predicted accuracy. Self-regulation interacted with condition, but the same interaction appeared on instructor-assigned grades, indicating pre-existing group differences. Perceived accuracy was unrelated to actual accuracy (R² = .001), so students' confidence cannot serve as a proxy for precision. Self-assessment accuracy was associated with final exam performance, although causal conclusions cannot be drawn. Grading conditions shaping self-assessment accuracy warrant systematic study.