Investigating adaptive decisional processes : effects of stimulus difficulty context

Until now, decisional research, and decisional models, have focused on response criterion, response bias, and stimulus discriminability as the sole factors that motivate decisional performance changes. The findings of the current study suggest that, in addition to these 'antagonists', the effects of stimulus difficulty context must also be viewed as a factor that induces systematic performance adaptation. A general review of the response time-probability literature provides a basis for the evaluation of recent decisional models in terms of their ability to account for known performance changes. Finally, two studies involving comparisons of horizontal visual extent were undertaken in order to quantify the effects of decisional context. In experiment 1, it was found that response times were longer, and accuracy and confidence higher, for 'target' pairs embedded in a difficult as compared with an easy context of remaining stimulus pairs. In experiment 2, after establishing baseline performance on 'target' pairs embedded in a relatively easy subset of remaining pairs, subjects were shifted to a context with a preponderance of very difficult remaining pairs. Response time increased and confidence decreased for the 'target' pairs for subjects working under a payoff scheme emphasizing accuracy. However, response time, confidence, and accuracy decreased following the context shift for subjects working under a strict 450 msec deadline. The results are discussed with respect to the predictions of Vickers' (1970;1979) adaptive model of decision; the one model found to be sufficiently comprehensive to describe dynamic and adaptive decisional processes.