The development of meaning contexts for empathic accuracy: Channel and sequence effects.

摘要

This study concerned contextual effects on empathic accuracy. Two broad classes of contextual cues in social perception were defined: immediate contextual knowledge, which is defined by all the local verbal and nonverbal cues that are available to the social perceiver; and cumulative contextual knowledge, which is defined by the overall history and evolving "meaning context" of an interaction or behavioral sequence.; Participants viewed thirty 15-second excerpts from three stimulus tapes, each of which portrayed a female client discussing her personal problems with a male client-centered therapist. The participants' task was to infer the specific thoughts and feelings of each client at thirty points during the tape. Immediate contextual knowledge was varied by manipulating the information channels in which the three stimulus tapes were presented to the participants. Participants were presented with either a full-channel audio plus video version of the interaction, an audio only version, or a video plus content-filtered audio version. Cumulative contextual knowledge was varied by manipulating the sequence in which the participants viewed the thirty 15-second segments. Half of the participants viewed the segments in the original order in which they occurred (cumulative meaning context preserved), whereas half viewed the segments in random order (cumulative meaning context disrupted).; The results revealed a very strong channel effect, with empathic accuracy in the video plus audio and audio only conditions being significantly (and comparably) higher than in the video plus filtered audio condition. The main effect of sequence was not statistically significant in an initial analysis. However, when the thought/feeling entries were categorized and analyzed according to their level of inferential difficulty, an intriguing and very strong interaction effect was found. For the easy-to-infer entries empathic accuracy was significantly higher in the original sequence condition, whereas for the difficult-to-infer entries empathic accuracy was significantly higher in the random sequence condition. Taken together, these findings suggest that empathic accuracy is dependent upon the verbal content of an interaction, and that when the judgmental task is difficult, the overall meaning context of the interaction may be misleading, leading perceivers to respond in a stereotyped fashion.