ABSTRACT
An experimental design was attempted to test the effectiveness of self-help group participation among people with severe and long-standing mental illnesses. A sample of 90 individuals was randomized to either experimental or control groups with experimentals invited to join self-help groups and given outreach to encourage participation. Control subjects were not invited to join self-help groups. Results showed low rates of 17% participation in the self-help groups for both experimental and control subjects. The small number of experimental subjects participating and the presence of crossovers from the control group yielded a sample too small for statistical analyses and the experiment was ended. Post hoc 316analyses comparing subjects who participated in self-help (15), nonparticipating subjects (75) and a comparison group of existing selfhelp group members (90) showed participants to have more severe psychiatric symptoms than either non-participants or members. Participating subjects were more likely to be female, though the selfhelp groups had more male members. Results indicate the need for multi-site studies in self-help group research.
