Harvard is used to finishing at the top of nearly every ranking of colleges.

But a survey in May's Yahoo! Life Magazine has placed the College a long way from even its third place finish in this year's U.S. News and World Report's annual college survey. The University was ranked 64th out of 100 schools in a survey of how well colleges use the Internet for educational purposes.

But those at Harvard were quick to dismiss the validity of the survey, which is based on interviews with students and administrators at 300 colleges.

"I think people need to read these surveys with a very large grain of salt," said Rick Osterberg '96, coordinator of residential computing support. "They made a big deal out of a very, very small number of indicators."

Dean of the College Harry R. Lewis '68 said he was skeptical of the methodology used in the survey.

"It seems to me rather silly that a school could get a high score by requiring a course in the Internet (which Harvard does not do) but get no credit for providing free Internet access to every student from the student's desk (which Harvard does and few other schools do)," Lewis wrote in an e-mail message.

Others cautioned that quantitative analysis of technology often overlooks the quality.

"Being able to count the number of courses which have Web pages may be a snazzy statistic, but it says nothing about how valuable these pages are to the pedagogy of the curriculum or how much they contribute either to the smooth administration of or the ideas and concepts central to a course," said Paul F. Bergen, the manager of the Instructional Computing Group (ICG), which assists the Faculty with academic computing.

In addition, a number of the fields in Harvard's ratings contained "N/A" responses, which lowered the school's ranking because no credit was given for incomplete or missing answers.

Despite having a large number of classes with Web pages, Harvard received no points in that area because no answer was given on the survey.

MIT ranked first in the survey, followed by Northwestern and Emerson College. RPI and Dartmouth rounded out the top five, with Princeton finishing 12, Yale 60 and Stanford 84.

The survey may not reflect Harvard's computing resources in several areas, officials said.

"It looks to me like the numbers they have for percentage of student and organizational Web pages is probably a bit low," said Daniel A. Lopez '97, former president of the Harvard Computer Society (HCS).

The survey results indicated that 10 percent of students and one-third of organizations had Web pages. But Lopez said the HCS server alone has nearly 200 organizations on its machine.

Osterberg said he thought the survey failed to ask about issues which have a greater impact on Internet use for educational purposes, such as the percentage of students connected to the network, and the computing support, online software packages, labs and networked laser printing available.

The survey ranked 35 factors in four categories. Forty-five percent of each institution's score was based specifically on academic usage, such as assignment of online homework and number of course Web pages.

Social uses and hardware each accounted for 22.5 percent and the remaining 10 percent was based on student services, such as the ability to register for courses online.

Despite the problems in the survey, some said such rankings would spur Harvard and other universities to focus on improving the use of the Internet for education.

"Judging universities on how well they are represented on the Web is a good thing since it will encourage us all to take more advantage of this important technology," said Professor of Government Gary King, a member of the Faculty Committee on Information Technology.

"But it is a bit like judging universities this Friday on how well their classes have incorporated the results from British elections held the day before," he added.

The results of the survey can be obtained online from http://www.zdnet.com/yil/content/college/100colleges.html