A session level analysis of the relationship between a group member’s fit with her group and PTSD symptom change in a sample of incarcerated women seeking trauma treatment

Abstract

The present study sought to apply the concept of person-group fit from the domain of organizational psychology to the domain of group psychotherapy. Specifically, using a time-series series design, the current study examined the relationship between an individual group member's fit with her group on two dimensions, perceptions of group climate and in-session intimate behaviors, and whether fit and standing on these dimensions were related to change in PTSD symptoms. An archival data set of 73 incarcerated women participating in six manualized (Trauma Recovery Empowerment Model [TREM], Harris, 1998) therapy groups for the treatment of co-morbid trauma and substance use disorders were analyzed. The relationship between the level of fit on these dimensions and change in PTSD symptoms as documented by participants' pre- and post-test scores on the PTSD Symptom Scale-Self Report (PSS-SR) was assessed. Using a session-level analysis (N = 1,606) and the Actor-Partner Interdependence Model (APIM) (Kenny, Kashy, Manetti, Piero, & Livi, 2002) both individual (Actor) and group (Partner) effects were modeled in order to test hypotheses about the individual, the group, and the fit between actor and partner and PTSD symptom change. Twenty-two of 73 women did not complete treatment. Analyses revealed significant partner effects for group members who completed both pre and posttest PTSD measures (n = 51) and those who did not (n=22). Specifically, members who completed both measures were in groups in which the other members perceived higher levels of engagement and lower levels of conflict. Results indicated that for both the individual and the other group members (partners), perceptions of the level of group engagement increased over time, perceptions of group conflict and avoidance decreased over time, and the average level of intimate behaviors in which group members engaged did not significantly change over time. PTSD symptoms decreased significantly between pre and posttest, however, no significant relationship was observed between fit of a group member and her group and PTSD symptom change. Results, limitations, and alternatives for data conceptualization and future analysis are discussed.