The influence of own and other outcome on satisfaction and choice of task difficulty

Hypotheses are stated about the relationships among the level of task difficulty chosen by a subject in a sequential task, expectancy of success, own outcome and that of a paired subject, and satisfaction with own outcome. Multiple regression analyses were used to test these hypotheses on data from an experiment in which subjects worked in pairs for six trials. It was found that the difficulty level chosen on a given trial was predicted by the difficulty level, the subject's outcome score, and the subject's satisfaction on the previous trial; that level of satisfaction on a trial was predicted by the subject's own outcome score and the difference in outcome scores between the two subjects on that trial; and that expectancy was predicted by difficulty level on the same trial. These results underscore the important mediating role of outcome evaluation in the control of behavior, and are consistent with previous research showing a dependence of evaluation on social comparison processes. Also consistent with previous research is the observed trend over trials to more difficult tasks. The “slope” of this trend depended on the initial difficulty level, a result which is readily accounted for by a simplified version of our hypotheses.