NCJ Number
205235
Journal
Violence and Victims Volume: 18 Issue: 6 Dated: December 2003 Pages: 641-657
Date Published
December 2003
Length
17 pages
Annotation
This study examined the internal structure of the Index of Spouse Abuse (ISA) in a sample of 583 low-income African-American women.
Abstract
Critical to the study of social problems is the issue of valid measurement. Researchers must know with certainty that the phenomenon they intend to measure is actually measured by the assessment tool used. Furthermore, the assessment tool needs to be flexible enough to correspond with evolving theories about violence against women. The current study assessed the internal validity of the ISA, which is one of numerous instruments available to study violence or abuse against women. The ISA was selected for validation because of its frequent use and its limited validation data. The ISA contains 30 items organized into subscales that measure 2 forms of abuse: physical (11 items) and nonphysical (19 items). The study focused on whether the subscales were conceptually clear. Data for the validation study were drawn from three related studies in which the ISA had been used. A sample of all African-American women (n = 583) was extracted from the complete dataset for analysis in the current study. In addition to the ISA, basic demographic data were collected from participants using the Demographic Data Questionnaire. Confirmatory factor analyses were performed on three models: (1) Model 1 tested the fit of Hudson’s and McIntosh’s original Index of Spouse Abuse two factor model; (2) Model 2 tested the fit of Campbell, Campbell, Parker, and Ryan’s three factor model; and (3) Model 3 tested the fit of an alternative model of physical and nonphysical abuse. Model 3, which was the theoretically specified alternative model, provided the best fit of all three models. The alternative model was informed by Tolman’s conceptualization of psychological abuse as a construct comprised of both controlling and emotionally abusive behaviors. Findings support the continued use of ISA, provided that proposed modifications are made. Modifications involve Tolman’s distinction between psychological abuse as behaviors that are meant to ridicule, intimidate, or patronize, and controlling abuse as behaviors that are meant to control another individual’s behavior. Tables, note, references