Francesco Bilotta

Bocconi University · Economics · 2026–27 job market

Fields

Behavioral & Experimental, Econometrics, Theory

Job market paper

Imperfect Metacognition in Standardized Tests

This paper argues that standardized tests measure not only what students know, but also how well they recognize what they know and do not know: their metacognition. I develop a framework that reveals metacognition from question-selection choices. When answering is costly, metacognitive students should omit the questions they are least likely to answer correctly, thereby improving their accuracy. Using data from a large field experiment, I find that students fall short of this benchmark. Students select only slightly better than random on average; selection quality varies widely, and a sizable minority select adversely. To explain these patterns, I estimate a structural model. Heterogeneity in metacognition accounts for a substantial share of score variation as well as gender differences in selection behavior: male students are better at recognizing what they know, whereas female students are better at recognizing what they do not know. Formula scoring—the standard rule used to discourage guessing—rewards only the former skill, so this asymmetry helps explain why gender score gaps widen with ability, a puzzle for traditional accounts of test-taking. More generally, the results show how test design determines which skills scores reward.