A critical survey of methodology used in empirical research on the social support, social network, and personality characteristics of abused women
Date
1990
Authors
Cross, Jay Evans
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis analyzes sixty-one empirical research studies concerning the social support, social network, and personality characteristics of abused women. These studies were located through an extensive search of the woman abuse published between 1969 and 1989. Using the MAP (Methodological Architectural Plan), each study was systematically examined for data sources and generation procedures, research design, reliability and validity of instruments, and statistical methods and measures.
The survey identified several serious problems in the research. Much of the social support and social network literature uses small, unrepresentative samples of shelter residents. In addition, the group "battered women" has been treated as an undifferentiated mass, which has reduced the validity of causal inferences. Further, in over half of the cases, battered women were asked to provide social support and social network information for their current out-of-abuse situation, rather than for the previous abusive situations, thereby minimizing any observable changes that may have occurred as a result of the abuse. The personality profile studies were also problematic; apart from the small clinically based samples of abused women collected, nearly all were cross-sectionally designed. This has led many researchers to conclude improperly that specific schizophrenic-type personality factors of the abused woman solicited and maintained the abuse, rather than such characteristics having emerged either from the abuse itself or as an outcome of the social isolation that the women may have endured. The implications of these findings for future abuse research are discussed.