Skip to content

Abstract Search

Injuries/Violence

Factor structure and concurrent validity of the severity of violence against women scale in zambia and south africa Amy Zheng* Amy Zheng Anthony J. Rosellini Jeremy C. Kane Kristina Metz Srishti Sardana Ross Greener Sithabile Mngadi Pertunia Manganye Lawrence C. Long Donald M. Thea Laura Murray Matthew Fox Mwamba Mwenge Saphira Munthali-Mulemba Sophie Pascoe

Background: The Severity of Violence Against Women Scale (SVAWS), is a 46-item questionnaire developed and validated among women in the US to assess two subscales: threatening acts and physical violence. However, it has not been evaluated using gold standard scale validation approaches and little is known about its psychometric properties in other populations and cultures.

Methods: Exploratory factor analysis (including parallel analysis) was used to evaluate the structural validity and psychometric properties of SVAWS among women reporting intimate partner violence in the past year in Zambia (N=246) and South Africa (N=368). Factor loading patterns were evaluated based on their strength and conceptual interpretation (e.g., eliminating items with loadings<0.3 or large cross-loadings). Cronbach’s alpha was used to evaluate the internal consistency. Concurrent validity analyses were conducted to determine whether constructs assessed by SVAWS were associated with depression, trauma, and substance use.

Results: Our findings in Zambia and South Africa indicated several non-salient factor loadings. While the original scale identified two subscales, our findings identified three underlying dimensions with good internal consistency: Threatening (α=0.90-0.96), Physical (α=0.86-0.87), and Sexual (α=0.81-0.87). In support of their concurrent validity, the three subscales were more strongly associated with trauma (Harvard Trauma Questionnaire r=0.32-0.51) and depression (Center for Epidemiologic Studies Depression Scale r=0.33-0.44) than ever reporting substance use (Alcohol, Smoking, and Substance Involvement Screening Test r=-0.03-0.28).

Conclusion: While SVAWS is an appropriate tool for assessing intimate partner violence in South Africa and Zambia, the underlying constructs may differ across cultures and require revisions to SVAWS items and highlight the importance of validating instruments in new contexts to ensure concepts are measured accurately and reduce bias.