An Agent Based Model of Disease Diffusion in the Context of Heterogeneous Sexual Motivation

Abstract

This project focused on building and analyzing an agent-based model of disease diffusion in order to explore the hypothesis that the relative risk associated with an individual's "sexual motivation profile" (SMP) is influenced by the distribution of strategies represented in the population - that is, that sexual motivation functions as a frequency dependent trait. Sexual motivation is hypothesized to be composed of a sexual inhibition system (SIS) and a sexual excitation system (SES), following the Dual Control Model of Sexual Response. Results of the model show that the relative risk of a SMP does vary depending on the relative representation within a population, but that that variance is constrained by agents' absolute values of SIS and SES. The model produced several parallels with empirical data on humans, suggesting that the model accurately reproduced some aspects of human sexual behavior. For example, agents' SES was a better predictor than SIS of total number of partners, while SIS was a better predictor than SES of Age at Infection. Also, the more accurately the agent population matched the human population, the more the model produced human-like results. Future work should focus on increasing the verisimilitude of agents and their environments, in order to make models more practical for designing and testing intervention and policy strategies.