66 DONG,DAI,AND WYER been given a two aforementioned options and After the synchro in the case ofo how much freedom they had hen they do the exerc the Results egend o her ootep observers(s.224. rticipant role Analyses of these data revealed main situational reactance(B=-55.p<01)significantly predicted th pa re to oth 9590 o nonsignificance (B08.p37).whereas the effecto esting that the effect of choice freedom on mity wa reater dispo preferences under free choice confirm this con- red and averaged with their estimate of pere eived threa ingle index Insummary.the results of this experiment indicate that the 72) ns (M 322.SD d bo cho by affecting their psychological reactance. signi Experiment 3 ed the Our interpretation of the effect of actors'synchronous behavior and observers,both freedom of choice (B=29.p<01)and Table 2 heit behavior to others'behavior ind dently of its relevance to Effect of Natre of Choicen Actors'and Observers Subsequent Conformity Tendency:Experiment 2 Choice fredom of ynchrom or reas ving synch Role ac nd sho ce synchromy 89 19978欲 e als reich,2006;see also Haidt,207 Haidt,Seder.Kesebir,2008)to do the synchronous behavior task or were told that the actors had been given a choice of the two aforementioned options and had chosen to perform the exercises. After the synchronous exercise, all participants completed the same product preference task employed in Experiment 1. In addi￾tion, they indicated how much threat the synchronous activities had posed to their behavioral freedom and how much freedom they experienced when they were doing the synchronous exercises (or, in the case of observers, how much freedom they had experienced when they watched the actors do the exercises). Finally, they indicated the extent to which the actors were forced to perform the synchrony task. All responses were made along a scale from 1 (not at all) to 5 (very much). Results Manipulation check. Analyses of perceptions of synchrony as a function of choice freedom and participant role indicated that the actors indeed had greater freedom of choice in free choice conditions than in forced choice conditions, F(1, 113) 122.20, p  .001, 2 .52, and this difference did not significantly depend on whether participants were actors (4.07 vs. 2.55), F(1, 113) 48.96, p  .001, or observers (4.10 vs. 2.24, F(1, 113) 74.71, p  .001), F(1, 113) 1.25, p  .26. Conformity. Participants’ conformity to others’ preferences is summarized in Table 2 as a function of freedom of choice and participant role. Analyses of these data revealed main effects of both participant role, F(1, 113) 9.24, p .003, 2 .08, and choice freedom, F(1, 113) 10.77, p .001, 2 .09. Actors generally conformed more to others’ preferences (M 9.59, SD 1.48) than observers did (M 8.71, SD 1.71), confirming the results of Experiment 1. However, both actors and observers had a greater disposition to copy others’ preferences under free choice conditions (M 9.62, SD 1.44) than under forced choice conditions (M 8.68, SD 1.73). Furthermore, the interaction of choice freedom and participant role was not significant (F  1). Reactance. Participants’ estimate of perceived freedom was reverse scored and averaged with their estimate of perceived threat to form a single index of reactance. As we expected, participants perceived greater reactance in forced choice conditions (M 3.88, SD 0.72) than under free choice conditions (M 3.22, SD 0.78), F(1, 113) 22.87, p  .001, 2 .17. Neither the main effect of participant role nor its interaction with choice freedom was significant (ps  .22). Regression analyses confirmed the mediating effect of reactance on the impact of choice freedom on conformity. Pooled over actors and observers, both freedom of choice ( .29, p  .01) and situational reactance (.55, p  .001) significantly predicted conformity. When both factors were entered as the predictor of conformity, however, the effect of freedom of choice was reduced to nonsignificance ( .08, p .37), whereas the effect of situational reactance remained reliable (.52. p  .001), suggesting that the effect of choice freedom on conformity was mediated by reactance. Bootstrapping further confirmed this con￾clusion: Based on 1,000 samples, the 95% confidence interval ranged from 0.3803 to 1.1604, excluding 0. Separate analyses of actors and observers, summarized in Figure 2, confirm this con￾clusion. In summary, the results of this experiment indicate that the freedom of choice over performing synchronous behavior influ￾enced both actors’ and observers’ adoption of others’ preferences by affecting their psychological reactance. Experiment 3 Our interpretation of the effect of actors’ synchronous behavior on their later behavior assumes that the actors consciously perceive their actions as a means of attaining the goal of behaving synchro￾nously. However, one might speculate that actors’ conformity in synchronous behavior conditions is induced by the similarity of their behavior to others’ behavior independently of its relevance to the goal to which the behavior was directed. If this were so, however, inducing actors to match others’ behavior incidentally, for reasons that are unrelated to the goal of behaving synchro￾nously, should not activate a copying-others mindset and should not influence their conformity in later situations. Experiment 3 confirmed this hypothesis. We also evaluated another possibility. Engaging in synchronous behavior can give rise to a sense of group cohesiveness (Ehren￾reich, 2006; see also Haidt, 2007; Haidt, Seder, & Kesebir, 2008) Table 2 Effect of Nature of Choice on Actors’ and Observers’ Subsequent Conformity Tendency: Experiment 2 Role Choice freedom of synchrony Forced synchrony Free-choice synchrony Actor 9.10 (1.57)a 10.07 (1.22)b Observer 8.27 (1.80)a 9.17 (1.51)b,c Note. Cells with unlike subscripts differ at p  .05. Standard deviations are indicated in parentheses. (A) Actor Group (B) Observer Group Choice Freedom Situational Reactance Conformity Tendency -.45** -.57** .31** (.11ns) Choice Freedom Situational Reactance Conformity Tendency -.67** -.69** .27* (.04ns) Figure 2. Mediating effects of situational reactance on the effect of choice freedom on actors’ and observers’ conformity: Experiment 2. The values in parentheses indicate the regression coefficient of choice freedom on confor￾mity tendency controlling for situational reactance. ns nonsignificant. p  .05. p  .01. This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. 66 DONG, DAI, AND WYER