Beer Pong and…Study Habits? You May Learn More from Your College Roommate Than You Realize

The fall semester is starting at colleges across the country, and millions of students are about to embark on a social and academic journey, leaving home to pursue their education. Whether it’s their first semester or their fifth, emotions run high as parents and students alike confront various worries, but every college students has at least two thoughts in common: will I get along with my roommate and will I be able to excel academically? New research shows that these two concerns are, in fact, inextricably linked.

In “Peer effects in higher education: A look at heterogeneous impacts,” Professors Amanda Griffith and Kevin Rask study roommate pairs at two selective American colleges to investigate whether college peers affect each other’s academic performance. They find that such effects do exist, but their size depends on the measure of academic performance used. Likewise, certain types of students are more affected by their peers than are others. Griffith and Rask also suggest that college peers affect each other academically through general exchanges of habits and interactions, rather than by teaching each other specific knowledge or facts.

Griffith and Rusk contribute to a growing body of research examining peer effects in higher education. But the importance of understanding peer effects extends far beyond the college setting, as researchers have long tried to understand how people impact each others’ academic achievement and, more generally, each other’s behavior and habits—from the preschool classroom and onward. Understanding those effects could have important implications for how classrooms are organized and how academic interventions are targeted.

Previous research on college peer effects has found mixed results, and Griffith and Rusk aim to reconcile those findings by using the same data sets to examine the different methodologies used. They also set out to identify the mechanisms that allow such peer effects to take place. Using 13 years of data from a medium-sized liberal arts college and four years of data from a medium-sized university, the authors are able to test the effects randomly assigned roommates have on each other’s academic outcomes. They use two outcome measures—regular GPA and a standardized GPA that accounts for different grading intensity across different courses—and are able to control for race, nationality, gender, aid status, and any effects that might be associated with a certain year or dorm building assignment.

The authors find evidence of peer effects, but the size of the effects differs based on the measure used. Interestingly, roommates’ SAT scores appeared to have no effect on future peers’ achievement, but a compiled measure of achievement in high school was positively associated with college peers’ GPA. Male students, minority students, and student receiving financial aid appeared to be more influenced by peer effects than their counterparts. Peer effects also appeared to be stronger at the liberal arts college than at the university. The authors find no evidence that higher-achieving peers are harmed by being paired with a lower-achieving peer, while the lower-achieving students show academic benefits.

To understand the ways that peers affect each other, Griffith and Rusk break down their data to analyze the academic achievement of roommates who take classes together versus those who do not. They find that roommates do not benefit academically from enrolling in the same classes. Roommates who are more socially alike and have similar habits, which the authors measure by whether the roommates live together for a second year, also do not have particularly strong academic effects on each other’s achievement. The authors conclude that students most directly affect their peers when lower-achieving students pick up beneficial habits or study strategies from higher-achieving students; peers do not affect each other’s achievement by directly exchanging specific academic information, like studying together for a test.

Griffith and Rusk conclude that certain types of students could benefit greatly from being paired with high-achieving peers, with no disadvantage to the higher-achievers. Further, their research highlights some of the mechanisms through which roommates’ social and academic lives intersect, and how those might differ at different types of institutions. Such findings suggest low-cost strategies that could have a significant impact on college students’ academic achievement. While this study focuses on the college setting, understanding how peers interact more generally would help our ability to intervene with struggling students and to identify how different types of knowledge are transmitted through groups. More research should be done to examine similar effects in earlier grades and throughout students’ academic careers.

Article source: “Peer effects in higher education: A look at heterogeneous impacts”. Griffith, Amanda L. & Kevin N. Rask, Economics of Education Review, 39(2014), 65-77.

Feature photo: cc/(peculiarnothings)