Nearly three years late, and amid speculation and some controversy, the National Research Council (NRC) will release its third Assessment of Research Doctorate Programs on Sept. 28. The report, viewed as the gold standard of graduate school rankings, will assess 4,838 programs in 62 fields at 212 institutions, including Cornell, as recorded during the 2005-06 academic year.

The rankings are important to major research institutions like Cornell because they help identify areas for improvement in graduate education. In the 2005-06 school year, Cornell had 3,240 doctoral students, comprising 16 percent of the student body.

The last time the NRC, which is part of the National Academy of Sciences, released a study ranking national research doctorate programs was in 1995. In that survey -- based entirely on reputation as judged by the nation's faculty members -- Cornell's graduate fields ranked among the very best, tying with Yale University for sixth place overall, according to an analysis of the data. Among Cornell's 36 doctoral fields reviewed by the NRC, 19 were placed in the top 10 among the universities surveyed. (What the NRC refers to as programs Cornell refers to as graduate fields.)

Fifteen years later (almost three years after the study's scheduled release), the NRC is using a markedly different methodology from that used in its two earlier studies in 1995 and 1982. Instead of giving programs a specific numbered ranking (for example, in 1995 Cornell's astronomy department was ranked fifth-best in the country), the new rankings will be based on ranges derived from 20 variables for each program as well as on subjective assessments.

"The value of the NRC study may lie less in the specific rankings and more in the ability to compare programs across multiple variables to understand how to improve graduate study," says Cornell Vice Provost Barbara A. Knuth, who is also dean of the Cornell Graduate School. "Cornell's NRC rankings should be viewed within this context."

She notes that the latest NRC data are essentially "a snapshot" of the 2005-06 school year. However, she says, "graduate study at Cornell has changed since the data were collected." Some examples of these changes include a 17 percent faculty turnover and a 5 percent increase in university funding of graduate students -- from both assistantships and fellowships -- between fall 2005 and fall 2008.

Knuth also points to the planning for and completion of several facilities in support of research, teaching and graduate education on the Cornell campus over the past five years. Among them: Weill Hall, incorporating the Joan and Sanford I. Weill Institute for Cell and Molecular Biology, which opened in 2008; the Physical Sciences Building, which is nearing completion; and the College of Architecture, Art and Planning's Paul Milstein Hall, which is under construction. None of these buildings, she notes, will be reflected in the NRC data.

In the 1995 NRC survey, Cornell's ratings were the strongest in the arts and humanities, as well as in engineering and in physical sciences, where most of the programs were ranked in the top 10. Cornell did not fare as well in the social sciences. However, says Knuth, since 2005-06 Cornell has increased support for the social sciences. The university's Social Sciences Initiative is strengthening interdisciplinary collaboration by expanding opportunities for researchers and students to work together through the Institute for the Social Sciences.

The 1995 survey was based on the evaluation of doctoral programs for faculty quality and program effectiveness by 8,000 university faculty members across the country. Scoring was recorded for each school in each discipline, and the universities then were ranked within those disciplines.

However, the rating of the top 10 schools was calculated not by the NRC but by the media, by adding up the number of programs a school had ranked in the top 10 in the survey. This time around, however, it might not be as simple to calculate the top 10 schools. That's because the new NRC report will contain two separate rankings of programs. One will display broad ranges for program rankings overall, spreading them across 20 key variables, from the percentage of faculty with grants and the median time to earn a degree to the percentage of female and minority faculty and of international students. The second will be based on survey respondents' subjective assessments.

The Cornell data for the 2005-06 academic year were obtained from central university records, then submitted to the graduate fields for interpretation and verification before being released by the Graduate School to the NRC.

The results of the NRC survey will be available to the Cornell community at http://www.gradschool.cornell.edu/nrc.