Predicting the Vulnerability of Women to Intimate Partner Violence in South Africa: Evidence from Tree-based Machine Learning Techniques

Abstract

Intimate partner violence (IPV) is a pervasive social challenge with severe health and demographic consequences. Global statistics indicate that more than a third of women have experienced IPV at some point in their lives. In South Africa, IPV is considered a significant contributor to the country’s broader problem with violence and a leading cause of femicide. Consequently, IPV has been the major focus of legislation and research across different disciplines. The present article aims to contribute to the growing scholarly literature by predicting factors that are associated with the risk of experiencing IPV. We used the 2016 South African Demographic and Health Survey dataset and restricted our analysis to 1,816 ever-married women who had complete information on the variables that were used to generate IPV. Prior research has mainly used regression analysis to identify correlates of IPV; however, while regression analysis can test a priori specified effects, it cannot capture unspecified inter-relationship across factors. To address this limitation, we opted for machine learning methods, which identify hidden and complex patterns and relationships in the data. Our results indicate that the fear of the husband is the most critical factor in determining the experience of IPV. In other words, the risk of IPV in South Africa is associated more with the husband or partner’s characteristics than the woman’s. The models developed in this study can be used to develop interventions by different stakeholders such as social workers, policymakers, and or other interested partners.

Identification of Associated Factors and Prediction for the Level of Intimate Partner Violence against Women in Sri Lanka

“…It indicated that the FNN provided a better prediction on the Violence Index than the GR model. This agrees with the conclusion of the study by Amusa et al (2020) who have shown that compared to the regression model, different neural network models were more preferable to predict the vulnerability of women to IPV. Another study by Babcock & Cooper (2019) stated the same conclusion for prediction on the history of arrest among men who were perpetrators of IPV.…”

Prediction on Domestic Violence in Bangladesh during the COVID-19 Outbreak Using Machine Learning Methods

“…The main objective of this paper is to address the growing academic literature by monitoring framework connected with the hazard of IPV exposure. The research aimed for machine learning approaches that understand concealed and dynamic data trends and regularities [31]. Through current review thus aims to establish a predictive method that is clinically applicable.…”

“Measure What Matters”: The Challenges Brought by Non-Categorisation of Gender-Based Violence and Femicide as a Crime at Higher Education Institutions, South Africa

“…Canli and Kaya (2016: 229), report that the lack of care and support from police and safety officers has contributed to the distrust in the government because of the misconceived and misplaced belief that sexuality is a personal and private matter that should not be spoken of in public and that a woman's sexual experience is shameful, whether chosen or not (Canli & Kaya, 2016: 229. The SAPS are said to not be interested in ensuring that proper services are constantly accessible to victims of GBVF (Lopes & Stone, 2018;Kaur & Ahuja, 2019). There is continued criticism that the quality of services rendered to women and children in the SAPS is inadequate and unsatisfactory (Gibbs et al, 2018;Canli & Kaya, 2016;Amusa et al, 2020). Victims have expressed dissatisfaction with the police, mainly because the police are slow to respond when GBVF incidents are reported (SAHRC, 2018: 13).…”