An exploration of whether the extent and orientation of the discrepancy in perceived and actual mathematical efficacy affects mathematical anxiety

A questionnaire survey was used to study the effect of the discrepancy between perceived and actual mathematical efficacy (discrepancy in mathematical efficacy) on mathematics anxiety, and the moderating effect of bias in mathematical efficacy estimation. A total of 582 grade 1 students in seniors high school in Nanning with an average age of 16 and 489 grade 2 students in seniors high school with an average age of 17 were selected. The results confirmed that (1) mathematics anxiety was significantly negatively correlated with mathematics efficacy, and significantly positively correlated with greater discrepancy between perceived and actual mathematics efficacy. (2) Variation in mathematics efficacy was much greater, given underestimation rather than overestimation of efficacy. Further, mathematics anxiety was significantly greater, given underestimation vs. overestimation of efficacy. (3) Discrepancy in perceived vs. actual mathematical efficacy positively predicted mathematics anxiety, and was moderated by bias in the estimation of mathematical ability. In the first-grade 1 students, mathematics anxiety was predicted by underestimation but not overestimation of efficacy.