Houses should realize identity, diversity

Housing, Dining and Residence Life has gotten down to the brass tacks of the house model. This week, HDRL announced its plan to randomly place residentially unaffiliated sophomores and juniors into 40 unaffiliated houses across West and Central campuses in Fall 2012.

A successful placement structure must reconcile the house model’s two objectives: helping students build strong communities and encouraging them to learn from one another’s differences. Community and diversity are not mutually exclusive goals—the new placement model must allow houses to scope for identity without insulating them from diversity.

Randomization is the best placement strategy in the new model’s inaugural year. As pointed out by Duke Student Government President Pete Schork, the success of the model hinges on houses developing unique personalities. Randomization in the first year will generate diversity as well any other method, and it is swift and simple to boot. Compounded with highly effective, student-driven social programming, randomized placement gives houses a fair shot at creating their own identities.

Sustaining these house cultures will be more difficult. After the new house model’s first year, absolute randomization is a poor placement strategy. The cultures that grow within the houses during a single year will be fragile. An influx of randomly selected incoming sophomores will fracture these nascent identities, destroying any gains from the initial randomization.

If randomization is the sole placement mechanism, then the house model will only recreate the current quad model. Blocks will be adjacent to one another, and there will be little incentive for the students within those blocks to reach out to their housemates. Especially after a year of living on East Campus, most sophomores will choose to nurture old friendships rather than to repeat the random residence hall experience of freshman year. No amount of programming, however lavish or well executed, will facilitate strong communities within houses with randomly assigned upperclassmen.

Any solution to this problem must blend randomized placement with placement based on student preference. Unaffiliated sophomores should be able to rank houses according to which houses values best reflect their own personalities. A computer algorithm ought to maximize preference satisfaction by placing as many students in their desired houses as possible. Unlike the recruitment process used by greek organizations and SLGs, unaffiliated students would not be judged on external factors like race, gender or class—only the objective fact of how well their preferences match up against the desires of others. Such a process renders students’ preferences equal and anonymous.

A computerized system would also allow HDRL to inject an element of randomness into house sorting if necessary. Should the preference-based system yield houses that are too homogeneous, loosening the parameters of preference adherence would immediately remedy the problem.

The new house model should not be a repeat of the East Campus experience. Total randomization that occurs twice in the residential life of a Duke student is excessive. To expect that the second randomization will yield communities resembling freshmen dorms is naïve and unrealistic. Instead, the new house model should embody a fresh and uniquely upperclassmen experience, one that encourages both common interests through house preference and diversity through randomization.