In 2018, Duke University stopped honoring roommate requests from incoming freshmen, an attempted corrective to the fact that first-year students increasingly picked their bunkmates online before even stepping foot on campus.
Officials said at the time that the change would give freshmen more opportunities to meet people from different backgrounds and worldviews. While pairings would take some lifestyle preferences into consideration, they would mostly be random.
An uproar soon followed: What if this put marginalized students in the position of having to teach their peers about their culture? Shouldn’t the dorm, above all, be a safe space?
In the end, the rollout went off largely without a hitch. Now, new research published in the Journal of Personality and Social Psychology suggests that Duke’s random-roommate policy could be a strategy for colleges to encourage interactions between students from different racial backgrounds.
The finding is timely: Getting students to talk to peers with varied perspectives is a high priority for college administrators at a time of inflamed tensions over the election and the war in Gaza.
The Duke study found that the chances of having a roommate from a different racial group increased with the random-assignment policy, and that having a roommate of another race was linked to a more diverse set of friends and more positive behavior in interactions with people from different racial backgrounds. The study did not find differences in relationship quality between roommates who were of the same race and those who were of different races. (Research on the topic isn’t uniform; in another study, published in 2020, Asian, Black, and multiracial students who chose their roommates perceived a healthier campus environment than their classmates of the same race who were assigned roommates).
The Chronicle spoke to Sarah Gaither, an associate professor of psychology and neuroscience at Duke and the study’s co-author, about how she got interested in studying roommate dynamics and whether the findings carry lessons for improving today’s campus climate. The conversation has been edited for length and clarity.
Where did the idea for your research come from?
I originally was introduced to this idea of studying college roommates when I was a graduate student at Tufts University. I ran a behavioral study in psychology looking at white first-year college students there: If they lived with a same-race versus another-race roommate, how does that shift their behavior at the end of their first year of college? And what we ended up finding was that if you lived with someone from another racial background, for these white students, when they came to our lab to meet a Black person they had never met before, they were way less anxious. The Black students also felt more comfortable. So over all, this experience of living with someone who’s different seemed to have this positive impact on behavior.
The most recent study was just published a couple of months ago. We attempted and replicated and extended the work I did at Tufts University here at Duke. So one thing we weren’t able to measure during my previous work was how roommate contact affects students of color in addition to white students.
Can you describe what the top-line findings were?
Our main finding was that living with a roommate of another race actually significantly expanded people’s social networks. So their friends were significantly much more diverse. And this was a finding that replicated across both white students and students from racial or ethnic-minority backgrounds, which was pretty exciting to see.
Not everyone was happy when Duke made its decision to randomize roommate assignments in 2018. What kind of pushback can colleges expect to receive, and how can they respond to it?
Duke’s not the first or the only university with randomized roommate policies. Johns Hopkins, Dartmouth, a number of universities have switched in recent years to randomized roommate policies. The main pushback is primarily from students from underrepresented backgrounds wanting to live within safe spaces, which is a completely acceptable con to thinking about these randomized roommate policies. That was the majority of feedback we got. We also got some feedback from students and parents, since it was announced after students had decided to enroll at Duke for the next year, that we were taking away student autonomy.
My argument, since I became this representative of the roommate-policy change at Duke — even though I had nothing to do with the policy, I was just assessing it — is that it’s not really universities’ role to “play God” and decide which identities matter more than others with roommate assignments. There’s more than just racial or ethnic-minority safe spaces that are needed on college campuses — sexual orientation, religion. All of these different identities could potentially create problems within any dorm environment. If a university is choosing to only match based on some identities and not others, it ends up creating a very unfavorable, inequitable environment. So in my view, a randomized roommate policy just for the first year of college, not their entire time, is the most equitable way to give all students a chance to experience the diversity and the new experiences they’re supposed to have when they start college for the first time.
Situate us in the time in which Duke made its decision to randomize roommate pairings. Can you tell us about how people were meeting on social media and what the pitfalls were?
Duke’s previous policy was you could choose a roommate or be randomly assigned. Most students were using things like Facebook, even social-media platforms such as TikTok, to find people in different groups with similar ideologies, similar interests. And they would find people and just start texting through group chats and things of that sort. Almost every university creates some sort of GroupMe or Facebook group for the incoming class or the incoming cohort. A lot of students will try and roommate shop there.
The other thing that we see a lot is that when students have someone else from their high school also going to the same college, they’ll choose their high-school best friend to move in with. Which seems like a really safe option. But … when you choose to move in with your best friend, you might not actually be the best match for living with them.
The other thing that our roommate policy does: It helps even the playing field. Now students don’t have any expectations about who this person is going to be. You don’t have to become best friends with them. Whereas if you are moving in with your best friend, you have different stakes on the table.
Some college administrators worry that political divisiveness will lead students to stick with peers of similar backgrounds. At an especially tense time on college campuses, what else should colleges be doing to encourage relationships across differences?
Every college campus in the United States has multiple affinity groups of all different types of student abilities, backgrounds, cultures. Having public showcases where students can be exposed to different dance styles from different groups, or food festivals, for people from different cultural backgrounds, can be a strategy.
As it relates to political differences, since this is an election semester across the U.S., tensions are really high. A random-roommate policy helps create these environments for students to practice talking about differing viewpoints. Without that practice, when you meet someone new for the first time who thinks differently than you do, you might say the wrong thing or do the wrong thing. But if universities can actually create contexts where things are designed and developed, with experts sitting on both sides of a dialogue for a current debate topic, it can give students tools, words that they can use and actions that they can take to help create less of a division between political-group memberships too.
