A causal model of the effects of feedback on self-evaluations, self-efficacy, goals and effort

Abstract

A causal model was developed to overcome weaknesses in Campion and Lords Control Systems model (1982) which proposed that people who experience a discrepancy between their previous goals and performance can achieve congruence in the future either by increasing their effort or by reducing their goals, but failed to specify when each of these responses would occur. A new model which integrates Bandura and Cervone’s (1983) concept of self-evaluative satisfaction or dissatisfaction arising from performance feedback and two dimensions of their concept of self-efficacy (confidence in task-specific skills and goal expectancies) was developed to resolve weaknesses in Campion and Lord’s discrepancy reduction explanation by predicting when people who experience discrepancy increase effort rather than reduce goals, and when those who achieve congruence increase effort rather than maintain goals and effort. The plausibility of the new model was tested in a longitudinal field study in which 503 students who enroled in a first year psychology course answered six questionnaires and completed five multi-choice tests as part of their normal course assessment. Path analyses offered some support for the new model but revealed a weak link between effort and performance and raised questions about the effect of self-evaluations on skills self-efficacy, goals and expectancies. Further analyses also cast doubt on the tenability of the discrepancy explanation for feedback functioning by showing that performance had more effect on subsequent responses than the discrepancy between previous goals and performance. In addition, other results showed that some responses depend on perceptions of the accuracy of the feedback. Two revisions of the new model were devised, both of which included previous performance instead of discrepancy as the only exogenous variable and focused on predicting effort rather than subsequent performance. The plausibility of the revised models were compared, and the results of people who rated their feedback as accurate were compared with those who questioned the accuracy of their feedback to explore the effects of the perceived accuracy of performance feedback on responses. Results favoured the revision which predicted that people who are dissatisfied with their performance feedback are more likely to increase goals and effort if they experience their feedback as motivating, and more likely to have low skills self-efficacy if they experience their feedback as debilitating. Together, the results suggested that perceptions of the motivating effects of feedback were more useful than Bandura and Cervone's concept of self-efficacy in resolving weaknesses in Campion and Lord’s model. Results for people who rated their feedback as inaccurate also favoured the revision which emphasised the motivating and debilitating effects of feedback. However, the weaker support for both models among people who rated their feedback accuracy as low confirmed that perceptions of feedback accuracy influence responses. The results indicated that performance feedback generally had less effect on skills self-efficacy, goals and expectancies of goal attainment among people who doubted the accuracy of their feedback. However, poor performers were more likely to increase effort if they rated their feedback as inaccurate and those who were dissatisfied continued to experience their feedback as motivating. In contrast, dissatisfying feedback became progressively less motivating and more debilitating among people who rated their feedback as accurate. This pattern of results suggests that rejecting the accuracy of feedback may help ameliorate the potentially debilitating effects of negative feedback and serve a positive function for poor performers by helping them maintain motivation and effort, and discouraging them from withdrawing from the situation. However, the flaws and limitations to the key concepts from the original models, apparent changes in both accuracy groups’ responses to feedback over time, inconsistent results and weaknesses in the present research design and measures suggest that considerably more research is needed to identify all factors which influence responses to feedback. Issues which need to be clarified in order to develop a comprehensive framework to explain and predict the effects of feedback on motivation and performance are discussed.