STATISTICAL MODELING OF FACTORS ASSOCIATED WITH SOME FORMS OF GENDER-BASED VIOLENCE AMONG WOMEN IN SELECTED STATES IN NIGERIA

Abstract

Gender-based violence (GBV) data in Nigeria has been modeled with one-part models. The distribution of male to female gender-based violence in Nigeria is rightly skewed. This suggests that models which can accommodate over-dispersion are imperative to investigate the factors influencing episodes of GBV. Therefore, this study used the negative binomial model to examine factors associated with number of episodes of male to female gender-based violence among women in Nigeria. A sample of 1298 women of reproductive age (15-49 years) within households in communities was obtained from a large survey conducted in 2011using a multistage cluster random sampling procedure. Information extracted includes socio-economic and demographic characteristics. The number of episodes of sexual denial, beaten up, and verbal abuse were the outcome variables used to represent GBV while the explanatory variables include age, tribe, level of education, monthly incomes, religion, alcohol use, and smoking cigarette. Descriptive statistics used were frequency tables, mean, standard deviation kurtosis, skewness, and negative binomial regression were used to model GBV. The mean age of women was 33.3years (SD=11.1years). About 64% were married while 1.8% were divorced. The proportion of women who drank alcohol was 17.1% while 42.7% were educated. The proportion of women who responded from the Igbo tribe was slightly higher (38.9%) while women from the Hausa tribe were (13.3%). Those women who smoke cigarettes were lower than those who did not smoke a cigarette. The mean number of episodes of male to female GBV through sexual denial, “beaten up” and verbal abuse were 2.57 (SD=10.68), 2.13(SD=8.54), and 2.68(SD=6.29) respectively. The results of the negative binomial analysis for sexual denial indicated that women who are Igbo were about four times more likely to experience sexual denial (IRR=3.75, 95%CI= 1.13-12.42). The analysis further revealed that women who were from the eastern part of Nigeria have a higher risk of experiencing “beaten up” (IRR=6.07, 95%CI= 2.47-14.92). Likewise, the model indicated that the Igbo (IRR= 3.40, CI= 2.43-4.77) and the Hausa (IRR= 2.66, CI= 1.70-4.16) were about three times more likely to experience verbal abuse compared to Yoruba women. Negative binomial regression was able to detect the over-dispersion present in the data on gender-based violence. The analysis showed that age, tribe, and income are associated with the number of episodes of Gender-Based Violence among women of reproductive age in selected states in Nigeria.