Dynamics: An Application of the Survey Data‑Driven Agent‑Based Model for University Students

Abstract

College-aged students are one of the most vulnerable populations
to high-risk alcohol drinking behaviors that could cause them
consequences such as injury or sexual assault. An important factor that
may influence college students’ decision on alcohol drinking behavior
is socializing in certain contexts across the university environment. In
the present study, we aim to identify and better understand the ecological
conditions driving the dynamics of distribution of alcohol use among
college-aged students. To this end, a pilot survey study is conducted
to evaluate students’ movement patterns in different contexts across
the Arizona State University (ASU) campus, and the results are used to
develop an agent-based simulation model designed for examining the
role of environmental factors on development and maintenance of alcohol
drinking behavior by a representative sample of ASU students. The
proposed model, which resembles an approximate reaction–diffusion
model, accounts for movement of agents (i.e., diffusion) in various contexts
and alcohol drinking influences within those contexts (i.e., reaction)
via a Susceptible-Infected-Recovered type individual-based model. Of
the four most visited contexts at ASU Tempe campus—Library, Memorial
Union, Fitness Center, and Dorm—the context with the highest visiting
probability, Memorial Union, is the most influential and most sensitive
context (around sixteen times higher an impact of alcohol-related influences
than the other contexts) in terms of spreading alcohol drinking
behavior. Our findings highlight the crucial role of socialization in local
environments on the dynamics of students’ alcohol use as well as on the
long-term prediction of the college drinking prevalence.