Comparison of Sequential and Simultaneous Responding, Matrix, and Strategy Variables in a Prisoner's Dilemma Game

Abstract

This experiment systematically compared the effects of simultaneous responding and sequential responding by subjects, using a 2 x 4 x 2 factorial design. The four strategy conditions were 10% Cooperation, 90% Cooperation, Tit-for-Tat (response- matching with a one-trial lag), and Free-Play (two subjects actually paired). The two PD matrices had average expected payoffs of $.00 and $.025 per trial. Subjects were 160 college freshman women volunteers who served in groups of two or four and were paid for their participation. Overall cooperation was moderately high (56%). The significant findings were a strategy effect (10% Cooperation lowest, then 90% Cooperation, Free-Play, Tit-for-Tat highest), an increase in cooperation across the fifty trials, an interaction of trials with strategies, and an interaction of matrices with response procedures. The latter finding supports the experimental hypothesis that simultaneous and sequential responding do not have equivalent effects.