70 DONG,DAL.AND WYER come rela edness,F2. 05可 A2,3h alizng-others 14.75.p<.and of the for ed.Between-cell directing rbehavior rather well-known charitie relative to lesser kno 4.17).F(1 134)=14.65.p<001,and this was true regardless of nously or a sly as in other studies.Howeve the ere linke to their own were of th ed an ibe the expe as synch i55.SD=268.F1 1341=16 uld lead them to expe tance.much as observers e their disposition to was synchronous les tha actors'con ity in thes Method 3.61.pda o the mindset that actors had. ned rand of the As shown in th he first two ons o able 4 nent were assigned to t e role o r known charities hen their behaviot was synch ous thar or nonsynchronous exe under condi it was not In c observer donated less money ns identical to those employed in other studies. ne conditions than in ei ing body m were given c iptions and picture This ticipants withou being mentioned.In each ca nt confimmed the conclusions draw inearlier stud s well as judgm hat is.the effect of behaviorl med the brand choi tied to actors ants in actua outcomes. were larg at all to 5 (very much) expe directed to the instr tality of the actor to the at Results r than to the actor ing them to c rather than le to wel ous in synchronous bchavi 4000D06)hn 53)=11153 p<01,=.68.Furthermore artic nts wh Experiment 5 Experiment 4 showed that directing obser the goal of actors'synchronous behavior eliminated the reac- participants'product choices (computed as in Experiment 1)as a134) 10.65, p  .001, 2 .07, and behavior/outcome relat￾edness, F(2, 134) 3.17, p  .05, 2 .05. However, the interaction of these variables was also significant, F(2, 134) 14.75, p  .001, 2 .18, and of the form expected. Between-cell comparisons shown in Table 4 indicate that actors’ donations to well-known charities relative to lesser known ones were signifi￾cantly greater when their behavior was synchronous (Mdiff 3.71, SD 3.54) than when it was asynchronous (Mdiff 0.10, SD 4.17), F(1, 134) 14.65, p  .001, and this was true regardless of whether the observers’ outcomes were linked to their own out￾comes or not. Furthermore, when observers’ outcomes were inde￾pendent of those of actors, they conformed less when the actors’ behavior was synchronous (Mdiff 2.52, SD 4.24) than when it was asynchronous (Mdiff 1.55, SD 2.68), F(1, 134) 12.62, p  .01, replicating the results of Experiment 1. When the observ￾ers’ outcomes depended on the actors’ success, however, their conformity (Mdiff 1.91, SD 4.41) did not differ from their conformity when the actors’ behavior was synchronous (Mdiff 1.55, SD 2.68, F  1). In fact, their conformity was not significantly less than actors’ conformity in these conditions (Mdiff 3.12, SD 3.61, p  .20), suggesting that observers might have vicariously acquired a copying-others mindset similar to the mindset that actors had. Separate analyses of the money donated to well-known and lesser known charities provided further confirmation of these conclusions. As shown in the first two sections of Table 4, actors donated more money to well-known charities and less money to lesser known charities when their behavior was synchronous than when it was not. In contrast, observers donated less money to well-known charities and more money to lesser known ones in synchronous behavior/independent outcome conditions than in ei￾ther synchronous behavior/dependent outcome conditions or asyn￾chronous behavior conditions, whereas their donations in the latter two conditions did not significantly differ. Discussion This experiment confirmed the conclusions draw in earlier stud￾ies and showed that these conclusions generalize to actual behavior as well as judgments. That is, the effect of behavioral synchrony on actors’ donation decisions was primarily governed by a copying-others mindset. Moreover, when observers’ outcomes were not tied to actors’ outcomes, they donated less money to popular than to unpopular charities when the actors’ behavior was synchronous, suggesting that their donations were largely gov￾erned by the reactance they experienced. When their outcomes were linked to actors’ outcomes, however, their attention was directed to the instrumentality of the actors’ behavior to the at￾tainment of the goal to which it pertained rather than to the actors’ behavior per se. Consequently, the reactance they experienced was decreased, leading them to donate rather than less more to well￾known charities. Moreover, thinking about the instrumentality of the actors’ behavior to the goal may have led them to acquire a copying-others mindset without actually performing the behavior themselves. Experiment 5 Experiment 4 showed that directing observers’ attention to the goal of actors’ synchronous behavior eliminated the reac￾tance they experienced when they thought about the actors’ behavior per se, perhaps inducing a copying-others mindset similar to that of actors. Our conceptualization also implies that directing actors’ attention to their behavior rather than to the goal of this behavior should induce reactance similar to that experienced by observers. Experiment 5 confirmed this prediction. In some conditions, actors were told that their goal was to perform exercises either synchronously or asynchronously, as in other studies. However, they were given pictures of the movements to be performed and told to imagine themselves performing the exercises and to de￾scribe the experience they would have in writing. We expected that the writing task would stimulate them to think about the feelings they were likely to have when performing the exercises and thus would lead them to experience reactance, much as observers would. Thus, we expected them to decrease their disposition to copy others’ preferences in a later situation. Method One hundred ten Hong Kong undergraduates (44 male, Mage 19.84 years, SD 1.72) participated for course credit. Participants took part in groups of 10 –12 and were assigned randomly to conditions of a 2 (behavior: synchronous vs. asynchronous) 2 (experience: imagined vs. actual) between-subjects design. All participants in this experiment were assigned to the role of actor. Participants in actual-behavior conditions engaged in either synchronous exercises or nonsynchronous exercises under condi￾tions identical to those employed in other studies. In imagined￾behavior conditions, however, participants were told that they would be required to engage in an exercise involving body move￾ments and were given descriptions and pictures showing the move￾ments to be performed. In synchronous conditions, they were either told that they would be required to do these exercises in synchrony with other group members. In asynchronous conditions, they were simply told that they would do the exercises with other participants without synchrony being mentioned. In each case, participants were instructed to imagine the situation in which they would be doing the exercises and to elaborate the experiences in a few sentences. Participants in each condition then performed the brand choice task administered in Experiments 1–3. Then, participants in actual￾behavior conditions indicated the extent to which they felt that they had done the exercise synchronously along a scale from 1 (not at all) to 5 (very much). Results Manipulation checks. Participants who actually engaged in the exercise task reported that their behavior was more synchro￾nous in synchronous behavior (M 4.00, SD 0.63) than in asynchronous behavior conditions (M 1.97, SD 0.78), F(1, 53) 111.53, p  .001, 2 .68. Furthermore, participants who imagined performing synchronous behavior used an average of 3.24 synchrony-related terms to describe the behavior they ex￾pected to perform, whereas participants who imagined asynchro￾nous behavior used none. Choice behavior. Table 5 shows the mean conformity of participants’ product choices (computed as in Experiment 1) as a 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. 70 DONG, DAI, AND WYER