Two of a kind? Roommates, from many backgrounds, learn to deal

A night like any other for Ben Himowitz ’05 – playing poker with several friends in Greylock, enjoying a relatively sedate evening. The scene, though, was visible to any passerby who cared to look in, and one, in particular, did – Himowitz’s first-year roommate, Mike Cerrito ’05. Cerrito grabbed some chips and got right into the game, staying for several hours and taking all of Himowitz’s money.

“It was great that he saw us through the window and was cool enough to come in,” Himowitz said. “We hadn’t planned it at all, and most of the time we end up doing things, we haven’t necessarily planned them. We just end up hanging out in weird ways like that.”

At Williams, though, it may not be that weird at all.

“I always see my roommate from freshman year around,” said Devin Fitzgibbons ’04, who roomed with John Arendshorst ’04 in Fayerweather three years ago. “Of course, we’re on the same team, and we live in the same house now, too. But I see him in other places. There’s really no way I could shake him.”

New evidence, however, shows that roommates might not just be someone to play poker or rugby with. According to David Zimmerman, chair of the department of economics at the College, who you are roomed with also is statistically significant in predicting what your GPA will be.

Zimmerman recently published a study entitled “Peer Effects in Academic Outcomes: Evidence from a Natural Experiment” exploring the GPAs of Williams roommates from 1990-2001, using each individual’s SAT scores as a proxy for pre-collegiate academic achievement.

“If high-ability students can influence the learning of other students,” he said, “then it makes sense to try and admit those of high ability. If you’re trying to teach someone to be a tennis player, you throw them in a tennis camp with other excellent players. It’s the same idea.”

Zimmerman found peer effects to be statistically significant when students were separated into a mid-range SAT group in between the highest and lowest 15 percentiles. All effects were more strongly linked with verbal scores than math scores, while one of the most pronounced effects was that students in the middle of the distribution were more liable to have their grades suffer if their roommate was in the bottom 15 percent of the verbal range. Even more interestingly, a roommates’ verbal score was a more powerful predictor of a student’s GPA than that student’s own math score.

“It did seem to be the case as well,” Zimmerman said, “that entries where the verbal scores were lower tended to have lower GPAs as a whole.” The JA SAT scores, though, were found to have no importance.

Having a roommate, of course, is not simply an academic endeavor. For many, it is most importantly a social interaction.

There are 175 first-year doubles on campus, spread between the six frosh dormitories. At a bare minimum, then, at least 1400 Williams students out of any given four-year spread have had the pleasure of rooming with one of their peers. Factor in the numerous doubles available to upperclassmen as housing options, and the roommate question appears even more unavoidable.

First-years, however, are most affected by their placement in cohabitation. Sharing bedrooms with siblings – once an extremely common practice – is less a feature of domestic living today than it was 20 years ago. Given that Williams students generally draw from above-average socioeconomic backgrounds, it is safe to say that the first night of First Days may be many students’ first exposure to communal living.

It’s impossible to predict how the interactions of two people will play out, particularly when the two have never met before. Some might attempt to minimize contact with the stranger in their bedroom, while others might jump headlong into a friendship. It is safe to say, though, that roommates will affect one another in realms including, but not limited to the social, hygienic, nocturnal and scholastic.

Luckily for all involved, the Housing Office has a well-established system to help alleviate the many problems that could arise from such an arrangement. Included in the substantial form packet given to pre-frosh is the “Housing Preference Form,” a seven-question info-gathering tool that helps to give the College an idea of the type of person they will be assigning. These parameters include taste in cleanliness, music and sleeping habits. A section of the form that allowed pre-frosh to indicate preference for a specific dorm was eliminated four years ago.

The actual room assignments are completed in early summer, with Housing linking up students using the above criteria. The task, usually carried out by Housing Services coordinator Linda Brown, was this year given to Community Life Coordinator (CLC) Brian Schwartz. Schwartz, like the other CLCs, was present in June, and offered to relieve some of Brown’s workload during the busy period.

“It’s pretty much hand-by-hand matching,” Schwartz said. “There’s a lot more single requests than there are space, so we have to put them together. It takes a while.”

Brown passed down the formula for assignments to Schwartz, including the relative weights to the different questions. Queries B (affinity for visitors) and G (sleep patterns) are considered more important than C and D (both dealing with musical taste).

“The entries themselves are pretty randomly put together,” Schwartz said. “My job was to avoid too much engineering. There were lots of overlaps, every response on the form was checked at some point, so it was pretty easy to keep kids with someone they could get along with.”

Get along they do. Housing estimates that almost a third of first-year roommates choose to live together again, suggesting that something positive is occurring within these entries. And when roommates do pick in together as rising sophomores, it can often be the basis for a lasting friendship.

“Nothing can be perfect, but it was amazing,” said Andrea Nogales ’04, who lived with her first-year roommate – Rachel Samuelson ’04 – for her sophomore and junior years. “I have a very large family, but that sort of makes me more protective of my personal space. With Rachel, it was always comfortable; I don’t know how they paired us so well.”

Nogales did have one complaint, however – “Everything was perfect except the clippity-clop of her damn clogs on the floor.”

Some Williams students, however, were downright critical of the peer effects of their counterpart. “My roommate was very smart and very lazy,” said Alex Smith ’06, who roomed with Dan Burns ’06 last year in Faye. “We had a class together, and one time I came back from four hours of studying for an exam in that class for the next day.”

Burns’ answer to Smith’s query as to whether or not he had studied for the exam was “What test?” “He was just sitting there playing video games,” Smith said. “I studied with him for like an hour before I went to bed, and then he ended up getting a better grade than me.”

Look on the bright side, Alex – at least he didn’t take your money, too.