Self-efficacy does not appear to mediate training’s effect on performance based on the moderation-of-process design

ABSTRACT Researchers in the area of training and motivation have concluded that training causes improvements in performance partially via training’s influence on self-efficacy. A set of studies employing a moderation-of-process design contradicts this conclusion. Specifically, when two groups of trainees did not know their performance was increasing during training, their self-efficacy did not increase. Two other groups were able to observe improving performance, which enhanced self-efficacy. However, both groups in both studies exhibited equal improvement in performance. This effect was shown in two contexts: training on a physical task and training on a cognitive task. These results demonstrate that self-efficacy might not always play a causal role in determining training’s influence. The studies highlight the value of a rarely used moderation-of-process method for assessing mediation.