This study investigated changes in male and female students' prediction and postdiction calibration accuracy and bias scores, and the predictive effects of explanatory styles on these variables beyond gender. Seventy undergraduate students rated their confidence in performance before and after a 40-item exam. There was an improvement in students' bias between prediction and postdiction judgments, although male students' decrease in bias was greater than female students. Female students' accuracy remained stable from prediction to postdiction, and male students exhibited greater accuracy then females. Additionally, the task centered and student centered testing styles positively predicted students' accuracy while the social centered style was a negative predictor, beyond the effect of gender. In addition, no attribution was a significant predictor of bias.