When it's an error to mirror...

Mimicry and imitation can facilitate cultural learning, maintenance of culture, and group cohesion, and individuals must competently select the appropriate models and actions to imitate. Mimicry and imitation also play an important role in dyadic social interactions. People mimic their partners’ mannerisms, which increases rapport and the partners’ liking of the mimickersA collaboration between psychologists and philosophers at the Univ. of California, San Diego asks whether and how mimicry unconsciously influences evaluations made by third-party observers of dyadic interactions. Their results indicate that third-party observers make judgments about individuals’ competence on the basis of their decisions concerning whether and whom to mimic. Contrary to the notion that mimicry is uniformly beneficial to the mimicker, people who mimicked an unfriendly model were rated as less competent than nonmimics. Thus, a positive reputation depends not only on the ability to mimic, but also on the ability to discriminate when not to mimic. Here is their experimental setup (click on figure to enlarge):

Figure: Illustration of the experimental paradigm and experimental results. Subjects watched two videos, in each of which an interviewer (model) interacted with an interviewee. After each video, subjects rated the interviewee’s competence, trustworthiness, and likeability. For each subject, one video showed a mimicking interviewee, and the other showed a nonmimicking interviewee. In Experiment 1, video frames were uncropped, so subjects could see the interviewer; in Experiment 2, video frames were cropped, so subjects could not see the interviewer, and mimicry was obscured. The interviewer’s attitude varied between subjects; some subjects saw videos with a cordial interviewer, and other subjects saw videos with a condescending interviewer. The graph shows the difference in average competence ratings between the cordial- and condescending-model conditions as a function of whether or not the interviewee mimicked the interviewer, separately for Experiments 1 and 2. Error bars represent standard errors of the difference between conditions.