A Spatial Latent Gaussian Model for Intimate Partner Violence Against Men in Africa

Abstract

This study used data from the Demographic and Health Surveys to examine the prevalence and correlates of Intimate Partner Violence (IPV) against men in Africa. Nationally representative data from 12 countries was analysed using a Spatial Latent Gaussian Model to capture the linear and non-linear nature of covariates while accounting for spatial heterogeneity. The results showed that female perpetrated abuse was significantly associated with education, alcohol consumption, intergenerational cycle of violence, polygamy, wealth, and type of union. As far as age of perpetrator (female) or victim (male) and spousal age difference were concerned, they had a non-linear effect on IPV. Furthermore, the spatial effects highlighted lack of spatial autocorrelation between the prevalence of IPV as one moves from country to country.