The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 21.0°F | Overcast

Counteracting Imperfect Rankings: MIT Should Post School Statistics on the Web for Prospective Students

Naveen Sunkavally

In August, U.S. News & World Report released its annual college rankings. The Institute fared well, as always, tied for fourth with Stanford University but behind Harvard University, Princeton University, and Yale University. However, as always, another hailstorm of criticism has bombarded the magazine.

This year, Stanford has led the way in condemning the rankings. Though the university continues to provide information to the magazine, Stanford has gone to the extreme measure of setting up its own web page with its own statistics to counteract U.S. News & World Report. At the risk of looking like copycats, MIT might benefit from quietly following Stanford's lead.

Though perhaps an informative source of college information, U.S. News is not an ideal source of information. Several of its weights are questionable, and its rankings create a false sense of precision.

For instance, one of the greatest disputes over the rankings concerns a newly introduced element last year called "value added" that uses the difference between the predicted graduation rate of a college and the actual graduation rate to determine the value that a college education adds. MIT received a -8 this year, Stanford a -4, and Caltech a -14. Are we to deduce then that none of these colleges, which are also ranked in the top ten, add any value to a student's education? After last year's controversy, this year U.S. News has opted to rename the "value added" measure to "graduation rate performance," but the name change does not eliminate the basis for such a statistic in the first place.

One may perhaps overlook the "value added" statistic, since it is only worth five percent of the rankings scheme employed by U.S. News, but a more serious problem arises when we evaluate the "graduation rate" statistic, which accounts for 16% of the grade. Both MIT and Caltech have comparatively low graduation rates, 89 percent and 83 percent respectively. U.S. News believes that the graduation rate reflects how well a college is educating its students; but could it be that a college with a lower graduation rate is also offering a tougher curriculum for its students, that it offers greater educational value to its students?

Which interpretation is more valid, and why does U.S. News choose the former interpretation? In fact, if we are to compare Harvard and Caltech, which placed ninth in the rankings, we would observe that Caltech fares statistically better or as well as Harvard in every category except in graduation rate. Had U.S. News chosen to see graduation rate as an indicator of a school's rigor, Caltech would be first.

That one questionable statistic can mean the difference between first and ninth raises additional concerns. The U.S. News rankings create a false sense of precision. In baseball, the rankings are clear: almost always one team can be judged better than another by comparing records or through head-to-head battle. In addition, because teams and players change quickly, the person who is better one year might not necessarily be better the next year. But colleges are not like baseball teams. Colleges don't suddenly jump from first to third as Harvard did one year or from sixth to fourth as MIT did within the last year.

If every prospective college student accepted the rankings with a grain of salt and acknowledged that there may be faults with the rankings, perhaps colleges wouldn't have to react to the rankings. However, not all students take the rankings with a grain of salt. They accept them as the truth, without fully evaluating the underlying methods used to make the rankings.

That is why I believe that MIT should set up its own set of statistics on a web page linked off its admissions page so that prospective students can have an additional and unbiased resource. Like Stanford, MIT should continue to supply statistics to the magazine, but unlike Stanford, MIT should do so quietly, so as not to antagonize the magazine and create controversy.

What type of statistics should MIT put up? At least all the base statistics used by U.S. News. By base, I mean those statistics that are not open to interpretation. Simply posting the graduation rate or acceptance rate without attaching a ranking to it would suffice. In addition, MITshould not choose to display only those statistics favorable to itself. Stanford, for instance, has great pride in listing out how many Nobel Laureates and Pulitzer Prize winners it has but fails to supply a "alumni giving rate" on its web page, a statistic that hurt Stanford in the U.S. News rankings.

By implementing a thorough, unbiased, uninterpreted source of statistical information as a link off its admissions home page, MIT can offer another source of information to present a more comprehensive view of the school. Students can interpret the data for themselves, decide what data is more important to them, and, in the end, choose the right for college for them.