EVALUATION OF A HOMELESS MENTALLY-ILL OUTREACH PROGRAM - DIFFERENTIALSHORT-TERM EFFECTS

Previously published research on interventions for persons who are hom eless and mentally ill has exhibited marked limitations in attrition, sample sizes, generalizability and outcome measures. This report prese nts results from an outreach and linkage project wherein the research design has better addressed these limitations. Successful outcomes in terms of the number housed were documented. However, significant chang es in participant functioning levels were not. Multinomial logistic re gression indicated that three variables were significant predictors of 4 month residential setting: recruitment source (shelter, psychiatric hospital or community mental health agency), client functioning clust er type, and hours of service from the homeless project. The latter fi nding suggests that project interventions contributed to positive chan ges in clients' residences. Implications of the results for future ser vice and research efforts are discussed.