Summary
Rotter's hypothesis about the power of the negative sign of anticipated reinforcement to change psychological situations so as to elicit generalized expectancies leading to avoidant behavior in Ss was tested. In an ambiguous angle-matching task, one group of 13 10-year-old boys was told that they would earn points for each match correctly made; another 13 of their male classmates were told that they had 50 points at the start and would lose a point for each incorrect match made. Otherwise, the two situations were the same for all Ss. Point-losers took much longer to make their matches than point-gainers. The groups did not differ in their choice to repeat the task. A procedure emphasizing losing, rather than winning, elicited Ss’ generalized expectancies for the efficacy of cautious and avoidant behavior while performing the task, but failed to affect choice to quit the task altogether.