Using Seligman's "learned helplessness" paradigm, androgynous and feminine women (as defined by the Bem Sex Role Inventory) either succeeded (contingent feedback) or failed (non-contingent feedback) at a concept formation task and were provided with internal (ability, effort), external (task difficulty, luck), or no causal attributions for their performance. Then a second (criterion) concept formation task was administered. The attribution manipulation was found to have no effect on any of the dependent measures and failed to interact with the other factors. For feminine subjects, failure increased the trials necessary to reach criterion on the second task ("learned helplessness"), whereas success had no effect. The performance of androgynous subjects, on the other hand, was unaffected by failure but facilitated by success ("learned competence"). Androgynous subjects responded to success by attributing it to their ability, whereas failure was attributed to task difficulty. Feminine subjects, on the other hand, attributed success and failure about equally to these two factors. Finally, whereas androgynous subjects attributed their successes in life to their ability and their failures to task difficulty, feminine subjects made the opposite attributions. The results suggest that the androgynous woman is likely to be more resistant to "learned helplessness" than is the feminine women. (Author)