Abstract
Due to the popularity of Bandura's notion of self-efficacy a series of three studies was conducted that examined the relationship between reported self-efficacy and performance. In each of the studies thiiy undergraduate students were asked to manage a computer simulated furniture factory. The subjects' performance on the simulation was correlated with their reports of self-efficacy. Study I successfully replicated the typical statistically significant positive correlation obtained by Bandura and colleagues. Study II successfully manipulated the relationship between reported self-efficacy and performance, producing a statistically significant negative correlation. Study III successfully manipulated the relationship between reported self-efficacy and performance, producing a strong, statistically significant positive correlation. The results of these three studies were examined in terms of their implications regarding the research methodology typically employed when studying the relationship between reports of self-efficacy and performance.