Past research on the effectiveness of laboratory training (the use of small group methods of experienced-based learning to produce personal change in attitudes and behaviors) has been deficient both in experimental design and in the concept of change. College students (N=128) participated in a study which used a series of 11 measurements to assess the nature of change (alpha, linear or pre- post-change; and beta change, instrument reactivity or scale recalibration) during a 13-week laboratory training session in 15 interpersonal communication skills. Students completed weekly Likert scales on two dimensions of the 15 skills: how good are you at the skill?, and how much change do you desire? Forty-four comparison subjects who did not receive the training completed the same self-rating form during their first and eleventh class sessions. An analysis of the results showed significant linear trends for all 15 pairs of regression equations. A significant positive change occurred during the training period for 13 of the 15 skills on the how good dimension. On the desired change dimension a significant negative change occurred for each of the 15 skills. Subjects perceived that their ability to perform the skills increased with training. They also continued to value increasing their ability to perform the skills throughout the training program. (BL)