Abstract
The present study was inspired by finding that a survey of research papers cited as demonstrating the effectiveness of brief training in communication skills failed to reveal evidence of successful training in "empathic respond ing." Two aspects of the general training situation were experimentally manipulated in an attempt to produce higher levels of "empathic respond ing" following a brief training program: ( a) the definition of the task, as well as the training instructions and materials, were deliberately phrased in a manner continuous with ordinary discourse. This was an attempt to solve the "translation" problem. (b) The purpose of the task was described as demon strating competence in assimilating an unorthodox set of rules, and not as accepting the validity of the trainer's conception of "empathy" and "being helpful." The results of the experimental manipulations were very dramatic: the group asked to display competence produced more than twice as many "empathic responses" as the group asked to learn "empathic responding," as assessed by an independent rater using a standard scale. In addition, 70% of the former group's responses were rated "empathic." The implications of this study for the general theoretical problem in psychology of evaluating capa city from performance are discussed.