The accuracy of self-reported grades in school

In educational research, self-reported grades are often used as indicators of actual grades, although this practice is frequently criticized. Therefore, we compared self-reported grades in eleven school-subjects with those from the last report card in a sample of N = 540 students (class level: 10). The correlations of self-reported and actual grades were very high (r >= .90), slightly lower correlation coefficients in few subjects were mainly due to smaller variances. Mean differences indicated a small overestimation of self-reported grades (d <= 0.18). Students with lower achievements overestimated their grades more than students with higher achievements (eta(2) <= .09). This overestimation did not correlate substantially and consistently with conscientiousness, the importance of good grades, academic self-concept, and academic cheating. In terms of validity the correlations of self-reported grades with these variables did not differ substantially from correlations of actual grades with these variables. These results support the practice of using self-reported grades for research purposes.