Make Me a Match

Mid-century meeting: roommates in the Fifties. Duke University Archives

The night after graduating from high school, Duke freshman Brett Jeffries, of Chicago, was chatting on the phone with a friend she had met during a campus visit earlier in the spring when she realized that she had not yet completed her roommate assignment form, which was due by midnight

She rushed to the family computer, worried that she might miss the deadline, but was surprised to find that the survey consisted of just twelve questions, easily completed in less than five minutes. Among the questions were blanks to request a specific roommate, indicate a FOCUS program, or request wellness or arts community housing. (The FOCUS program brings together first-year students in groups to study specific intellectual themes; each FOCUS group is assigned to one freshman dorm.) But beyond that, there were just five basic questions: bedtime, wake-up time, whether they study with music, whether they smoke, and whether they mind a smoker as a roommate.

Looking back, Jeffries said she was optimistic about being assigned a roommate she would get along with, but still, "You spend five minutes putting down information, and based on that, you're going to end up getting paired up with someone for an entire year."

Those seemingly insignificant five minutes, a mere blip in the steady stream of preparations made by outgoing high-school seniors, allow staff members at Residence Life and Housing Services a crucial glimpse into the incoming class, at this stage, a box of puzzle pieces waiting to be properly aligned.

Still sitting in her Central Campus office at nine on a Tuesday evening, well into her fourth straight day of freshman roommate pairings, Jen Frank, coordinator of housing assignments, is energized. It's hard to tell where that energy comes from. There are no crumpled coffee cups, no candy wrappers spilling off her desk and onto the floor. The air conditioning is blasting to neutralize a stifling summer heat.

Her goal is to get enough done this week so that she can leave by five or six on Friday to celebrate her twenty-seventh birthday, which she says will officially mark her stepping into her "late-twenties." This young woman controls the fate of roughly 1,700 fresh-faced, seventeen- and eighteen-year-olds, whose introduction to dorm life at Duke may hinge on her decisions.

It is Frank's job to assign a roommate to every student on Duke's campus. It's a solitary job, and that's by necessity. Interruptions keep her from getting on a trademark roll--one of which kept her working until 1:30 on a Saturday morning last year.

She flips through a binder with floor plans of all the halls on East Campus, letting it fall open at the third floor of Gilbert-Addoms. Like most halls on East, this particular floor is divided between males and females. Frank diagrams the split using what she calls "the all-so-creative pink and blue" color scheme. Half of the floor is outlined in pink, half in blue. Some of the rooms in the middle can be shimmied back and forth to accommodate additional students of one sex, but "if I go too far down the hall," she explains, pointing to the division, "I'll overload one bathroom--and underload another."

Her next challenge is Pegram, the headquarters for the FOCUS program on the genomics revolution, and also home to this year's performing-arts community. On her computer screen are two open spreadsheets, roommate survey data for students in those two programs.

Different spreadsheets hold the names of students who answered the five questions in a given way--say YNNYN--and would thus make a good match. After students who require a specific building are divided accordingly, it's really a matter of matching Y's and N's. Relatively simple, but try doing it 800 times.

In a red binder, which lists all of the female bed spaces on campus, Frank opens to Pegram. Each room appears on the page with blanks representing open bedspaces. She flips forward to the doubles.

The first room is easy--one FOCUS student along with a self-selected roommate. She draws brackets around them so she knows not to split them up as she makes adjustments later. Another FOCUS student is easily paired with an arts student whose answers match perfectly. A third proves more difficult: Her answers don't quite match those of any arts student, and FOCUS students cannot be paired together--it's a requirement of the program that they be spread throughout the building. Frank leaves the space blank, preferring to seek a good match from the larger pool of freshmen. By 10 p.m., 592 of the incoming class of 1,759 have been assigned.

The actual process of roommate pairings is routine for housing administrators, Frank's colleague Don Love '78 had explained earlier. "But it's emotionally fraught for the students. In a way, it really will affect the next four years, and their whole lives. Our goal in this whole process is happy students." He pauses, then adds with a wry smile, "Yes, this is the Pollyanna moment of our conversation. But really, if they're happy, then we're happy."

That's a tough balance to strike. The "Campus Life and Learning Project," an ongoing study headed by Kenneth Spenner, a sociology professor, and Anita-Yvonne Bryant, coordinator of multicultural services for Counseling and Psychological Services, aims to trace friendship networks through the classes of 2005 and 2006. The researchers have found that more than 51 percent of freshmen surveyed in the spring of 2002 "did not like" their roommate. (Twenty-six percent said they get along but are not friends, and 23 percent were friends.)

"It's a bit more pessimistic distribution than I think most people would give you," Spenner says. However, "one of the principles of social networks is the notion of homophily: Birds of a feather flock together. The more you get away from that, to randomly assigned roommates, the more it's going to differ." That's not necessarily a bad thing, he adds. It just depends on what you are aiming for.

While a handful of colleges and universities go through a more elaborate process in an attempt to ensure a good friend match--Davidson College, for example, assesses potential roommates using the Myers-Briggs personality test--Alan Hargrave, president-elect of the Association of College and University Housing Officers-International, says that Duke's approach is the norm: A good match is a livable match, he says. Thus, the questions about smoking, bedtimes, and noise--the bare necessities.

To first-year students, on the other hand, a good match might connote something closer to a best friend, he says, which would suggest the need for questions about, for example, interests.

But Love argues that fostering a livable situation while leaving the details to chance may actually help to promote diversity--of ethnicity, interests, politics, and geography, for example--and the idea that "I can get along and live with someone different than me." It's not as if they are setting students up for conflict, reality-TV-style, but neither are they providing a dating service. And if conflicts do arise, the housing office is there to take care of them.

According to Frank, the stark statistics revealed by Spenner and Bryant's research may be evidence of another trend. "It used to be that we were dealing with a lot of people who had always had their own room to themselves. Now we're getting to the point where they've always had their own bathroom to themselves."

Frank and Love have made some serious changes geared toward improving matches since taking over last year. Based on the advice of colleagues at other schools, and disbelief at their own statistics, they added the answer choice "social smoker" to question four. Whereas last year's class had only thirteen self-identified smokers, this year's has six, but also forty-four social smokers. That distinction is likely to be appreciated by thirty-plus, nonsmoking, would-be-roommates.

They've also introduced an online form, giving students the opportunity to make changes up until the last minute. That advance was inspired by the classic story of students filling out their forms with parents looking over their shoulders. In the first week the form was open, Frank had already noticed several cases of students logging in a second or third time, minutes or days later, to change preferences.

Frank returns to her binders, now contemplating whether to match two unusual male names together. She doesn't want to bring them undesired attention, but ultimately decides that the pair is a keeper. "Their preferences match," she says. "Besides, this is not a class full of Johns and Marks. It's got a pretty diverse set of names."

She recalls how earlier in the week, she identified two pairs of males with identical names. The obvious joke, she says, would be to pair each of them with their match. However, she says she's strongly tempted to make two separate roommate pairs that consist of the same two names. She laughs. "It would be nice, in a slightly more subtle manner. That's assignments humor."