The relationship between self-efficacy and amount of mental effort invested in mathematics problem solving by adults

This exploratory study hypothesized an inverted U relationship between self-efficacy (Bandura, 1997) and mental effort during mathematics problem solving by adults following suggestions by Salomon (1984) and Clark (1998). It was hypothesized that both low- and high-efficacy adults would invest lower mental effort in mathematics problems. Low-efficacy adults were expected to avoid the task because of a failure expectation. High-efficacy adults were expected to use automated mathematics problem solving strategies and invest low amounts of mental effort. Medium-efficacy adults were hypothesized to invest the greatest amount of mental effort due to their belief that they would not be successful without investing mental effort to solve the problems.;Subjects were 48 adult volunteers (21 males and 27 females) ages 18--25 enrolled in a university introductory psychology course. Mental effort was measured using two methods found to be effective previously: self-report and a "dual task" requirement to estimate time passing during problem solving (Cennamo, 1992). Mathematics tasks with cognitive loads of 10 were used, following Sweller (1994). Subjects were given misleading performance feedback to encourage them toward low, medium, or high self-efficacy levels.;Results partially supported the hypothesis of a curvilinear, inverted U relationship between self-efficacy and mental effort. High-efficacy subjects reported significantly lower mental effort than medium- or low-efficacy subjects did. Yet the mental effort scores of low- and medium-efficacy subjects were similar, perhaps due to cultural differences affecting medium effort for Asian subjects (Mizokawa & Ryckman, 1990). Self-efficacy was only partially affected by the feedback manipulation.;Mathematics performance was significantly correlated with self-efficacy scores (r = .43). Scores for medium- and high-efficacy subjects were similar even though high-efficacy subjects invested less mental effort than medium-efficacy subjects did. This buttresses the argument that high-efficacy adults invest less effort in learning and problem solving because they have high levels of automated skills.;Subjects with high mathematics task self-efficacy had significantly higher time estimation performance than low-efficacy subjects, supporting the theory that people with high self-efficacy invest low amounts of mental effort. Subjects applied the mental effort they did not need to use in the mathematics task to the time estimation task.