Empathic Perspective Taking and the Situational Malleability of the Communal Self-concept

Abstract Previous research has suggested that communal self-descriptions might be less malleable than agentic ones. We propose that the communal self-concept is malleable as well, and that empathic perspective taking activates the communal self-concept. In two studies, participants watched a video and were either instructed to empathically take the perspective of the target person or to remain objective and detached. In Study 1, we used a repeated-measures design and found that participants' communal self-descriptions changed after empathic perspective taking, but not in the objective condition. In Study 2, we assessed the accessibility of the communal self-concept by measuring response latencies. Participants in the empathic perspective-taking condition responded faster for communal traits and judged more communal traits as self-descriptive than participants in the objective condition. Keywords: AgencyEmpathic perspective takingSelf-concept; CommunionSituational malleability Notes The complete script of both videos used in the present studies can be obtained from the first author. Originally the scales consisted of 10 items each, but we excluded items for which the corrected item–scale correlation was below .30 for both times of measurement (agency: creative; communion: sensitive, trustworthy). Excluding these participants enhanced the effect of the manipulation, F(1, 66) = 42.19, p < .001, ηp 2 = .39. If these participants are reassigned to the other experimental condition according to their answers to the manipulation check, the findings are basically the same. Pre-test communion scores differed slightly between conditions (somewhat higher in the objective condition). Therefore, it might be argued that a ceiling effect may have contributed to the present findings. We cannot fully rule out this possibility. However, considering both the means and standard deviations there was still room for changes in an upward direction. As agency scores also differed slightly between conditions, we assume that although we randomly assigned participants to the conditions there was some sampling error which we hoped to overcome with a somewhat larger sample in Study 2. This includes 15 male participants in the empathic perspective-taking condition and 16 in the objective condition. The original instruction in Polish was: "Wyobraź sobie, jak ta osoba się czuje!", which stands for "Imagine how that person feels" thus incorporating both affective and cognitive components of empathic perspective taking. Excluding these participants enhanced the effect of the manipulation, F(1, 81) = 24.82, p < .001, ηp 2 = .24. There were no other significant effects, all Fs < 1. If these participants are reassigned to the other experimental condition according to their answers to the manipulation check, the findings for the endorsement rates as well as for latencies are basically the same. To account for the different endorsement rates, we also analyzed the response latencies for "yes" answers only. The results were by and large the same. Although response latencies for the practice words and for communal trait words were significantly correlated, the correlation was only moderate in size, r = .45, p < .001, and variance inflation factors below 1.32 suggest that multicollinearity was not an issue in the regression analyses. Additional informationNotes on contributorsMirjam Uchronski This research was supported by a grant from the German Research Council to the second author (Ab 45/10–1) and by a grant from the German Alexander von Humboldt Foundation to Andrea Abele and to Bogdan Wojciszke. Parts of the research were presented at the 12th Meeting of the Social Psychology Section of the German Psychological Society, Luxembourg, 2009. We thank Tamara Hagmeier and Katarzyna Klos for acting in the videos. Furthermore, we thank Olga Bialobrzeska, Michal Parzuchowski, and Bogdan Wojciszke for their assistance and help in organizing and conducting the second study.