A one-year follow-up revealed that the positive effects of group counseling with delinquents were still present on three of four concepts as measured by the semantic differential technique. In addition, significantly more of the experimental group had been released and not recommitted to the institution. Implications and the need for specific research are identified.

a highly significant difference (t = 5.46, d j = 8, p < .OOl) in favor of the latter score. I n accordance with the interpretation of A-State(g), these two results indicate that Ss reported significantly greater feelings of tension, nervousness, worry, or apprehension during the 10-hour marathon experience than either immediately before or after the session. Furthermore, these findings give empirical support for the heightened intensity of feelings that have been described as characteristic of the marathon interaction(loP ll). It remains for further marathon therapy research to explore the extent of the relationship between a shift in anxiety levels and concomitant changes on other personality measures. SUMMARYThis study investigated the impact of marathon group therapy on state and trait anxiety. Nine university students responded to measures of state and trait anxiety immediately before and after participation in a 10-hour marathon group. The hypothesis that A-State would decline from pre-to posttherapy while A-Trait

Thirty counseling practicum students' perceptions of self as counselor, ideal counselor, and self as counselor as perceived by other practicum students were measured by the semantic differential technique before and after the supervised practicum experience. Posttest perceptions of self as counselor, ideal counselor, and ratings by other practicum group members were significantly closer to the hypothetical model counselor than were the pretest ratings.

Retired military enlisted personnel are more likely than retired officers to find their choice of a second career to be fairly commensurate with their military work duties since officers rarely move into a high civilian managerial position. Consequently, it was hypothesized that adjustment to retirement would be more successfully accomplished by enlisted men than by officers regardless of how well they planned or how long they had been retired. A three-factor analysis of variance for unequal groups was conducted to test the statistical significance of rank, number of years retired, and preretirement planning on retirement adjustment. The effects of all factors and all interactions of factors were found to be nonsignificant with the exception of preretirement planning (p <.001). Those who planned well adjusted well, regardless of rank or number of years retired.