Self-evaluation of social ability: An action identification analysis

A critical discrepancy in the self-evaluation literature concerns whether self-uncertainty leads to accurate or biased self-evaluations. After discussing the prior research which has rendered this discrepancy, I propose an experiment to reach a resolution to the discrepancy. Principles of action identification theory are used in proposing that the link between self-uncertainty and self-evaluation is moderated by the amount of evaluative threat inherent to the situation in which the self-evaluation is taking place. The experiment established that subjects who are induced to identify their actions in low level terms are more likely to exhibit biases when pursuing self-evaluations of their social ability, with the biases manifested in these subjects' preferences to be evaluated in non-diagnostic ways. The implications of this general finding and suggestions for further research are discussed.