Some limitations of the attributional learned helplessness model on understanding effects of (non-) contingency: A controlled study in Dutch adolescents

Abstract

This study aims at testing for the various aspects central to the attributional learned helplessness model in an experiment designed to optimize ecological validity. Two-hundred and seventy eight Dutch secondary school pupils participated in a learned helplessness experiment, held in the classroom. At pretest an IQ test, the ASQ and some other materials were filled out. Corresponding to random assignment to a noncontingent negative, a noncontingent positive and a contingent group, subjects received at post-test, six days later, bogus feedback about their IQ test result. At post-test also, a parallel form of the first IQ test and several other tests were presented, interspersed by ratings of expectancy, certainty and momentary mood (repeated measures design). Relating to (non-)contingency, the validity of the learned helplessness model seems to be limited: effects of (non-)contingency on expectancy, achievement and mood on the whole cannot be fully understood without considering interactions between subjects' characteristics, schooltype, and the type of the dependent variables. In relation to the alleged mediating functions of attributions, it is revealed to be of little importance which attribution is made, but it is important whether any attribution is made at all. In respect of the intensity of learned helplessness, results do not show much support for the hypothesized relationships. Discussion of results lead us to conclude, that learned helplessness theory may gain by supplementing it with other theoretical notions, such as those relating to sex differences, fear of success and the relationship between emotion and cognition.