A supervision analogue was used to determine whether counselor trainees' preference for supervision method affects the outcome of supervision. Subjects were enrolled in a counselor education master's program. Two treatment groups of 8 subjects received supervision according to their preference; 2 groups of 8 did not. A control group of 8 subjects received no supervision. Supervisory sessions focused on increasing levels of empathic understanding. An analysis of empathy ratings indicated that learning empathy was not contingent upon receiving the preferred mode of supervision. Type of supervision received was a significant factor. Subjects receiving didactic supervision were rated higher in levels of empathy than subjects in other groups. Interaction effect between supervisor and supervision method was also significant. (Author)