Predicting disposition following brief residence at a shelter for battered women

ArticleinAmerican Journal of Community Psychology 9(5):559-566 · January 1981with2 Reads
Impact Factor: 1.74 · DOI: 10.1007/BF00896476


    Despite the recent proliferation of community shelters designed specifically to serve battered women, remarkably few evaluation studies have been conducted to determine the subsequent disposition of women following brief residence at these agencies. In particular, whether or not a woman returns to live with her assailant has far-reaching implications for the type of counseling services she requires as a shelter resident. From data available at admission, the current study employs discriminant function analysis for maximizing prediction of a battered woman's living arrangements 6 weeks following discharge. Results of the analysis improve the prediction of living arrangements from a 52°7o rate of accuracy using base rates alone to an 80% rate of accuracy using discriminant function coefficients. Implications for the development of more effective service delivery and for future research are discussed. It has been only recently that domestic violence and wife-abuse in particular have been recognized as a social problem of epidemic proportions. In the United States in any one year, approximately 1.8 million wives are beaten by their husbands. Nearly 30°70 of all couples report experiencing a violent incident at least once in the course of their marriage; estimates of actual incidence of physical abuse in marriages run closer to 50 to 60°70 (Straus, 1977-78). By 1978, some 100 shelters had been established as a refuge for battered women, and their number has grown steadily since (McShane, 1979).