University Times

A new ranking of universities by research performance has Pitt at No. 53 internationally and at No. 38 among all U.S. universities.

Pitt is tied for 53rd with Osaka University, Japan, in the 2003 “Academic Ranking of World Universities,” a first-ever ranking from the Institute of Higher Education at Shanghai Jiao Tong University, China.

According to information on the Shanghai Institute’s web site (http://ed.sjtu.edu.cn/ranking.htm), the rankings attempt to measure “the academic ranking of universities worldwide by evaluating their academic or research performance” based on five performance indicators.

Those indicators are: the number of Nobel Laureates affiliated with the university, weighted in favor of recently awarded Nobel prizes; the number of “highly cited” researchers in the years 1981-1999; the number of articles published by faculty in Nature and Science during 2000-2002; the number of articles published by faculty in the Science Citation Index Expanded and the Social Science Citation Index, and a composite of academic performance per faculty member.

Unlike companion ratings, such as the annual rankings of U.S. News & World Report, the emphasis in the Shanghai study is solely on research without regard to teaching or professional school quality. U.S. News also includes indicators based on faculty resources, entering students’ test scores, student graduation rates, faculty/student ratios and the subjective rankings of peers’ perceptions and student selectivity.

Each indicator in the Shanghai Institute’s rankings counted for 20 percent of the total score. The highest scoring institution was assigned a score of 100; other institutions were calculated as a percentage of that.

In Pitt’s case, the total score was 33.0. That broke down to 0.0 in the Nobel category; 37.0 in the highly cited category; 26.8 in the Nature and Science bracket; 65.1 in the science and social science indices category, and 25.5 in the score per faculty member.

Steven Husted, Arts and Sciences associate dean for Graduate studies and research at Pitt, said rankings sometimes reflect behind-the-sciences agendas of preparers who choose criteria that make their own institutions look good. “But, from my point of view, this is a very credible and important measure,” Husted said. “Clearly, citations in academia matter. They demonstrate not only that the research is published but that it’s being read and is having an impact on other researchers in the field.”

The Shanghai Institute published the top 500 institutions overall, while ranking only the top 100 in order. U.S. universities dominated the overall rankings, with Harvard at the top, followed by Stanford, California Institute of Technology and the University of California at Berkeley. Britain’s Cambridge (at No. 5) and Oxford (No. 9) were the only two non-American institutions that made the top 10. Tokyo University was the highest rated Asian institution at No. 19.

In addition to Pitt, Pennsylvania institutions that made the top 100 internationally included Penn (No. 18), Penn State (tied at No. 40 with the University of Utrecht in the Netherlands, and the University of Southern California) and Carnegie Mellon (tied at No. 61 with Rice).

Some of the five indicator categories carried restrictions. For example, the institute counted the number of Nobel laureates in physics, chemistry, medicine and economics between 1911 and 2002 by the prize winner’s affiliation. Different weights were set for every 10 years: the weight of 100 percent for laureates in 2001-2002, 90 percent for laureates in 1991-2000, 80 percent for laureates in 1981-1990, and so on, ending with 10 percent for laureates in 1911-1920.

The rankings took into account the number of highly cited researchers in 21 broad subjects in the life sciences, medicine, physical sciences, engineering and social sciences.

For articles published in Nature and Science, the institute weighted the order of author affiliation. A weight of 100 percent was assigned for first author affiliation, 50 percent for second author affiliation, 25 percent for third author affiliation, and 10 percent for other author affiliation.

Academic performance per faculty was defined as total scores in the other indicator categories divided by the number of full-time-equivalent faculty.

According to the institute’s web site, the distribution of data for each indicator was examined for any significant distorting effect and standard statistical techniques were used to adjust the indicator if necessary.

—Peter Hart