“…Put differently, in a class of 100 students, the student with the 50 th rank always has the median performance, regardless of whether the class as a whole is weak or strong. Our measure of overplacement is commonly used in research on confidence (e.g., C. Anderson et al, 2012;Belmi, Neale, Reiff, & Ulfe, 2019;Ehrlinger, Mitchum, & Dweck, 2016;Emich, 2014;Friehe & Pannenberg, 2019;Muthukrishna et al, 2018). To calculate discrepancy between self-estimated placement and actual placement, we simply subtracted actual from estimated placement; but for tests of covariation involving overplacement, we used the residuals when regressing selfestimated placement on actual placement, which capture aspects of beliefs that cannot be explained by true performance, consistent with existing approaches (C. Anderson et al, 2012;Cronbach & Furby, 1970;Dubois, 1957;John & Robins, 1994a; see Supplemental Materials for expanded discussion on calculating discrepancy).…”