Date of Award


Document Type

Open Access

Degree Name

Bachelor of Science



First Advisor

Roger Hoerl




slavery, common, abolition, coercive, contemporary


Mathematical models are essential for mapping the future spread of HIV/AIDS and for evaluating the impact of prevention efforts. These models are often relied on by policymakers to make decisions about where to target resources and it is crucial that the model predictions are as accurate as possible. One model that is relied on to predict HIV Incidence is the UNAIDS Modes of Transmission model (MOT.) A major limitation of this model is that it does not incorporate different heterosexual sex risk factors, such as the presence of violence. In this thesis I created a revised MoT model in South Africa, which incorporates a concerning sexual risk factor: the presence of violence. South Africa was chosen because of its high rates of HIV prevalence and gender based violence. The results of the revised MoT model in South Africa indicate that gender based violence is a significant contributor to the HIV/AIDS epidemic. With our current parameter estimates for violence, the model demonstrates that the violence sub groups both relatively account for a greater proportion of new infections than the non-violence groups and on an absolute level, the violence sub groups have a significantly higher HIV incidence rate than the non-violence groups. For example, the violence sub group of individuals engaging in casual heterosexual sex is predicted to account for 7.18% of the total new infections in South Africa, where as the non-violence group is estimated to account for 2.82% of the new infections. In absolute terms, the violence group is expected to have an HIV incidence rate of 2.097% whereas the non-violence group is predicted to have an HIV incidence rate of .498%. In the case that violence was entirely eliminated, the model predicts that overall HIV incidence in South Africa would drop from 1.9% to 1.22%. The designed experiment sensitivity analysis similarly demonstrates that sexual risk behaviors associated with violence are significant drivers of HIV incidence. Based on these results, I suggest that addressing gender based violence be an important goal of the HIV/AIDS response.