Abstract
25 postgraduate students were exposed to a control and 4 self-consequation conditions in a Latin square design. The experimental tasks were arithmetic problems matched for difficulty. Following each experimental condition, participants responded to a questionnaire to ascertain subjective differences in perceived valence of the particular type of consequation. This procedure yielded two groups designated as ‘adequate’ or ‘inadequate’ within each condition. Results of post hoc comparisons gave no predictable pattern between contrasted groups within conditions despite predicted between-condition differences. Factors other than perceived incentive valence of the consequences were responsible for the predicted differences between the self-consequation conditions.