Treatment Outcome in a Population of Homeless Veterans: Testing Models of Mediation and Moderation
MetadataShow full item record
Recent analysis indicates that veterans are overrepresented in homeless populations compared to the general public (Fargo, et al., 2012). This is a concerning statistic with implications of risk deserving empirical clarification. Unfortunately, a clear pattern of factors influencing homeless veterans’ success or failure in treatment programs has yet to emerge (Salvatore, Sussner, Smelson, Kline, & Losonczy, 2008). The current study undertakes the objective of contributing to such knowledge. Toward this end, descriptive characteristics of a sample of homeless veterans in treatment for homelessness in the Houston area are reported, binary logistic regression analyses examining predictors of successful treatment outcome is conducted, and tests of mediation and moderation completed. Results indicate treatment duration positively associates with treatment outcome, and mental illness partially mediates a negative association between institutional disaffiliation and treatment outcome. Discussion of results highlights implications for clinical efficacy as well as data driven theoretical conceptualizations of problems inherent to homelessness.