Success in college may take years of preparation, dedication, and hard work — but it also helps to have brainy roommates, a Stanford study suggests.
Sharique Hasan, an assistant professor of organizational behavior at Stanford Graduate School of Business, and Surendrakumar Bagde, an officer in India's Administrative Service, recently tracked the academic records of approximately 2,000 male and female students enrolled at a competitive new public engineering college in southern India.
It turns out that individuals randomly assigned to live with high-performing roommates, in large barracks-style dormitories, performed significantly better on tests at the end of the second semester than students placed with lower-achieving peers.
Furthermore, the so-called "roommate effect" was stronger than the influence exerted by friends and study partners. It also seemed to transcend social barriers; students benefitted from having really smart people in their dormitory rooms even when their castes were different.
The findings echo similar studies done in the United States. In his widely cited 1973 paper, "The Strength of Weak Ties," Stanford sociologist Mark Granovetter promoted the idea that novel bits of information (job leads, for example) tend to come from acquaintances rather than close friends moving in the same circles. More recently, economist Bruce Sacerdote found that randomly assigned roommates at Dartmouth affected each other's GPAs. Another study, by Cornell sociologist David R. Harris, showed that white students assigned a roommate of a different skin color wound up being more open-minded about race.
Hasan hopes his new study will be of particular use to education policy makers in India. With half of its population below the age of 25, the rapidly developing nation is gearing up to enroll an additional 25 million students in colleges and universities over the next 7 years, at an additional cost of nearly $200 billion, according to the Deloitte Global Education Newsletter. Nearly two dozen education bills were awaiting approval in India's parliament last year, most of them dealing with higher education.
"My grandmother lives in India, and I visit at least once a year, so I care a lot about the educational system there," says Hasan, who joined Stanford's business school faculty in 2010, after earning his doctorate at Carnegie Mellon University.
"This is probably the first study on peer effects in India," he adds. "In America we have been doing quantitative social science on peer effects since the 1970s, but in India there's just not a lot of quantitative information to infuse policy debates. Our goal is to give them more insight into what is going on." However, he cautions, "We will need many more studies before we can make very specific policy recommendations."
To conduct the research, Hasan and Bagde, a former doctoral colleague from Carnegie Mellon now serving in India's Ministry of Finance, followed students from a variety of geographical regions and castes who had been randomly assigned by the school to 92 same-sex dormitory rooms. Each large room accommodated 20 to 25 students; their beds were all in a row.
All of the students had taken the same multi-day, government-administered board exam for admission in the previous year. Once on campus, they all took the same first-year classes in math, chemistry, physics, and English.
The researchers surveyed the students early in the academic year about their social networks on campus. Then they compared the students' entrance scores with their subject test scores at the end of the second, third, and fourth semesters.
Predictably, the roommate effect was most pronounced after the first year, when all students were taking the same courses. "If a student is working on an assignment that he finds challenging, his peer may be a valuable resource who can clarify what the teacher is asking, help the student work through a difficult problem, or suggest additional study materials," the researchers noted. A diligent roommate also can be helpful indirectly by setting an example for the amount of time that should be spent studying and attending class.
In contrast, "chosen" peers — friends and study partners — had significantly less influence on academic achievement. "You would think that friends would have more of an impact than roommates," Hasan says, "but people tend to select friends and study partners who are similar to themselves. Apparently, it's that diversity of exposure from the overall environment that seems to have the greatest impact."
A related finding had to do with caste: India's ancient system of social stratification based on hereditary occupation. Although caste-based economic disparities have declined on the subcontinent in recent decades, Hasan and Bagde hypothesized that high-achieving students would relate better — and therefore be more influential — with students of their own caste. In fact, this didn't happen; the roommate effect was similar whether the interacting students were of the same caste or not.
In this case, the relaxed dormitory setting probably helped. "Outside the college context, individuals from different castes often live in different parts of town," the researchers noted. "Their parents are often employed in occupations with different status, and they may attend different schools." When the structural barriers are lower, as in college, they observed, "the interaction is significantly higher."
In future studies, the researchers hope to follow the same group of students to see how peers affect them as they attend graduate school, launch into careers, and build their own families.
Sharique Hasan is an assistant professor of organizational behavior at Stanford Graduate School of Business.