This study, the Principal Investigators first NIH award, examines in-session behaviors in drug abuse preventive intervention groups. The primary goal of this study is to initiate a program of research that will increase the knowledge-base about how clinicians influence within-session participant behaviors and group processes which, in turn, influence proximal outcomes ('Attendance', 'Parental Involvement'). The research plan in this study borrows established theories, methods, and procedures from the psychotherapy research literature to examine in-session behaviors in preventive intervention groups. Findings from this study will provide a foundation for translating results from prevention trials into immediate improvements in the delivery of effective interventions. Videotapes of group sessions will be coded (statement by statement) and rated (molar level assessment of the extent to which a behavior was present in a segment) to identify distinct classes of in-session behaviors. Confirmatory factor analyses will be conducted to test for the presence of clinician intervention ('Supportive', 'Directive', and 'Facilitative'), participant behavior ('Bonding to Clinician' and 'Bonding to Group'), and group process ('Cohesion') constructs. Hierarchical linear modeling will be used to examine the hypothesized relationships between clinician interventions, participant behaviors, group process, and proximal outcomes ('Attendance' and 'Parental Involvement'). Microsequential analyses will be used to elucidate the immediate (speech act level) relationships between in-session clinician interventions and participant behaviors.
