Theoretical and Empirical Investigations of the Dynamics of Homophily and its Impact on Students' Achievement, Decisions, and Well-Being

Stanford University, Stanford CA

Investigators

Abstract

People are more likely to interact with individuals like themselves. One word to describe this is ‘homophily’; when this is the case, people have biased social connections. This homophily affects people’s sources of information, opportunities, behaviors, and support. As a result, homophily may have important social effects. This project will use game theory and data analysis to study homophily and its consequences. The data portion will track the evolution of social connections over time, primarily in school settings. The project will examine (i) how homophily evolves and changes as students spend more time in a given setting and (ii) trace the impact of that homophily on students’ psychological well-being, achievements, and decisions about course selection and fields of study. The team will also use mathematical techniques to build modes of (i) how friendships and homophily patterns evolve over time and (ii) models of how homophily and social learning help and hinder people in learning about the potential benefits of specific decisions. The results will help us learn how social interactions aid and hinder both students and the broader society. The project will provide new data, facts, models, and understandings of the dynamic evolution of homophily, how it differs across traits, and how it depends on traits such as risk aversion and empathy. This includes work on how friendship and study partner networks evolve among students. The team plans to test causal hypotheses using data on student interaction patterns, perceived well-being, stress levels, and academic achievement. The new theory will focus on network formation and the evolution of homophily over time. The first new model will differentiate between characteristics that are easily observed by agents, such as gender and ethnicity, and other that take repeated interactions to observe, such as specific personality characteristics. The second model will example the impact of homophily. It will consider how people make decisions based on information that they learn from friends’ experiences. The team wants to determine whether and how homophily exacerbates herding behavior, whether herding behavior can differ across groups, and whether groups can crowd each other out. Additional work will focus on how the efficiency and diversity of choices made by individuals depends on the level of homophily, the diversity of a group’s aptitudes across actions, and the riskiness of the various actions. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.