Willing or able? The meanings of self-efficacy

One hundred and twenty college students (97 women) completed measures of anticipated performance for a fear-based snake behavioral avoidance task (BAT) and a skill-based basket-shooting task. Half of the participants were instructed to identify the task steps they believed themselves able to do (standard self-efficacy instructions). The remaining participants were instructed to identify the task steps they would be willing to try. Prior to completing the anticipatory performance measures, half of the participants in each of the "able to do" or "willing to try" conditions received information, called disambiguation instructions, about Kirsch's (1982) hypothesis that people frequently use these terms interchangeably for fear-based tasks but not skill-based tasks. Results revealed that, for participants who did not receive the disambiguation instructions, participants endorsed a similar number of BAT steps regardless of questionnaire instructions. On the basket-shooting task, participants asked to identify steps they were willing to try endorsed a greater number of steps than participants asked to identify steps they were able to do. For participants who received the disambiguating instructions, participants asked to identify BAT steps they are able to do endorsed a significantly greater number of steps than participants asked to identify steps they would be willing to try. On the basket-shooting task, participants asked to identify steps they were willing to try endorsed a greater number of steps than participants asked to identify steps they were able to do. These results are consistent with Kirsch's hypothesis and challenge the construct validity of measures of self-efficacy for fear-based tasks. Implications for the assessment of self-efficacy in other behavioral domains are discussed.