For the first time ever, the College has released figures detailing the breakdown of ethnic group and athlete representation in blocking groups over the last three years.

The data--which Dean of the College Harry R. Lewis '68 presented yesterday morning at Committee on House Life meeting--suggests a decline in the number of "large" blocking groups that are composed of more than 50 percent of a particular ethnic group.

The College also looked at the representation of varsity athletes in blocking groups, finding that the percentage of large blocking groups in which varsity athletes constitute more than half of the members rose slightly in 1997 and more dramatically in 1998.

The data also shows an increase in the size of blocking groups, which may prompt the College to decrease the maximum number of people per blocking group in the future, according to many of those present.

College administrators and several House masters at the meeting said they were not surprised by the data, which were compiled to help investigate the concerns that several minority resident tutors voiced last spring, according to Associate Dean for the House System Thomas A. Dingman '68.

The numbers, they said, did not show strong enough evidence to support the criticisms of the tutors, who in an open letter to the administration criticized the policy to remove student choice from the upperclass housing process.

In the letter, they warned that randomization would strip minority students of the support networks they had once found in certain Houses in the era of ordered choice.

They also cautioned that randomization might encourage minority students to form blocking groups with large concentrations ofpeople from their own ethnic group.

According to the data, the percentage of groupswith a majority of Asian students fell from a highof 12 percent in 1996 to about 4.9 percent in1998. From a high of 56 percent in 1996, thepercentage of groups with a large concentration ofwhite students fell to slightly under 40 percentin 1998.

In addition, after an increase from 1996 to1997 in the percentage of groups with a majorityof black students, the figure fell to 8.5 percentin 1998.

The data includes information from blockinggroups between 1996 and 1998 and defined "large"blocking groups to be those with eight or morestudents.

Of those "large" groups, the College computedthe number of groups with at least one member ofthe specified ethnic or athlete group. From there,it calculated the percentage of those that drew 50percent or more of their members from thatparticular population.

Lewis acknowledged the College's difficulty inselecting the specific groups to include in itsmeasurement--settling finally on groups it namedas "Asian," "Afro-American" and "White," inaddition to "Athlete."

Many also cautioned that with only three years'worth of numbers, it is somewhat difficult to drawsignificant longterm conclusions.

Eliot House Resident Scholar Nicky Sheats, along-time resident tutor who helped organize theletter last year, said there are several otherfactors the data does not adequately reflect butwhich must be considered when reviewingrandomization.

Sheats said the decline in the percentage ofgroups with a large concentration of black orminority students may mask the fact that thosestudents are forming smaller but more ethnicallyhomogeneous groups.