Task performance influences general self-efficacy, even without increases in the skills required to achieve success.

The current article tests whether task performance influences general self-efficacy without increases in the skills required to achieve success. To do so, an experimental design is applied in which participants predict a random future event, and the relationship between prediction task performance and self-efficacy is observed. This article also tests whether this specific performance/self-efficacy relationship is moderated by (a) perceived illegitimacy of predicting the future and (b) self-assessed ability to predict the future. The results show that prediction task performance indeed influences general self-efficacy, and neither of these two moderators has a significant effect on this relationship. Therefore, performance on a low-stakes task does influence general self-efficacy-at least temporarily. More importantly, task performance is shown to influence self-efficacy even without increases in the skills required to achieve success.