Attributional style questioned:Psychometric evaluation of the ASQ in Dutch adolescents

This study presents data on the psychometric properties and validity of the Attributional Style Questionnaire (ASQ), administered as part of a learned helplessness experiment among 278 Dutch adolescents. Confirmatory factor analysis showed that there was only weak evidence for the presumed ASQ dimensions. Further analyses revealed that situations seem to explain more variance than attributional styles. Correlations with mood and depressive symptoms were low or absent. The ASQ unsatisfactorily predicted attributions for noncontingent success or failure on an IQ test. Experimental effects on mood, expectancy and achievement were poorly predicted by the ASQ. Predictive and discriminant validity of ASQ subscales were weak, results reflecting sex-dependency. In contradiction to the concept of depressive attributional style, internal and stable styles for good outcomes were associated with negative emotions following noncontingent success. Girls with these styles showed relatively more hostility after noncontingent success, suggesting that the success offended sex-related norms. Unless the test format causes the disappointing results, this study suggests that: (1) the concept of cross-situationally consistent traitlike attributional style should be questioned, (2) theories about effects of failure and success should take sex differences and situation specifity into account, and (3) the construct of conscious-verbal, explicit attribution as a determinant of emotions and behavior may be inadequat