摘要:College students were assigned to 2 groups of 4 participants each in a reversal design. On each trial participants chose individually how many tokens to bet, and then collectively chose a row on an 8x8 matrix with a plus or minus sign in each cell. After that the experimenter announced a column that determined whether the group won or lost the bets. Before the trial ended, participants had to distribute their earnings. In experimental condition A the group won in trials after distributing proceeds equally on the previous trial, and in condition B they won only after unequal distribution in the previous trial. Results show that the external contingency on distribution (or, as we suggest, metacontingency) selected the groups’ distribution of their earnings.