Modelling women’s education and intimate partner violence —Methodological challenges and pitfalls: Evidence from Demographic and Health Surveys in Central Africa

Intimate partner violence (IPV) is now recognized as a serious threat against girls and women, and it has health and socioeconomic consequences for women and the whole nations. As such, it has received increasing attention during the last three decades and the international political and research agenda. Scholars devoted much time to determine the correlates of intimate partner violence to devise more effective interventions, especially in sub-Saharan Africa recording higher prevalence of IPV. Despite these efforts, findings are still inconclusive partly due to lack of a unified approach in operationalizing IPV and modelling strategies. Using 2013-14 Demographic and Health Survey in the Democratic Republic of the Congo, this paper estimated the net effects of women’s education using different modelling strategies (linear regression, logistic regression, Poisson model and its variants, multivariate—linear/probit regression). Findings indicated that there is no clear evidence of the associations between women’s education even though some estimates reached statistical evidence (e.g., women’s education and emotional violence). Yet, previous studies found that higher education was protective against IPV. Indeed, as women’s education increased, the risks of emotional violence decreased. This was not evident for physical violence and sexual violence even though the estimates went in the expected direction. Scholars should weigh between various estimation techniques to provide robust findings which could be very helpful to devise more effective IPV—interventions in sub-Saharan Africa and worldwide.