Citation Analysis Reveals Top Canadian Universities In 20 Disciplines

Sidebars Editor's Note: The newsletter Science Watch-published by the Philadelphia-based Institute for Scientific Information (ISI)-last turned its attention toward Canada more than three years ago. The publication examined Quebec universities' contribution to the nation's overall performance in science (3:1; June/July 1992). In its November/December 1995 issue (6:10), Science Watch expanded its scope to cover universities throughout Canada. Using records of total citations and citation impact, Science Watch ranked the top three Canadian universities in 20 fields.

 Following is Science Watch's analysis, reprinted here with permission of the newsletter and ISI. For more information on the citation databases discussed in the article, contact Christopher King, editor of Science Watch, ISI, 3501 Market St., Philadelphia, Pa. 19104; (800) 523-1850, Ext. 1341. Fax: (215) 387-1266. E-mail: cking@isinet.com.

Just over a year ago Science Watch surveyed research at United States universities, with the top 100 federally funded schools vying in a "tournament" of top-10 rankings based on citation impact in 21 fields (Science Watch, 5[9]:1, October 1994; 5[10]:1, November/December 1994). Now, Science Watch turns its gaze northward, scrutinizing university research in Canada. Examining performance in 20 fields over the last five years, Science Watch ranked Canadian universities according to two measures: citation impact (average citations per paper) and total citations. The top three performers in each field during the period 1990 to 1994, as measured by impact, appear in the top table. The bottom table shows the rankings in each field according to total citations.

Two institutions in particular-the University of Toronto and the University of British Columbia-clearly showed their clout in both tables, appearing in the rankings far more frequently than any of the other institutions. Still, other institutions asserted their leadership in select fields, especially through their high citations-per-paper marks. To obtain these rankings, Science Watch drew data from the Institute for Scientific Information (ISI)'s Canadian University Indicators on Diskette, 1981-1994, a database containing publication and citation statistics on 45 Canadian universities in nearly two dozen fields.

Leveling The Field

Different citation measurements, of course, will produce different results. Tabulating an institution's total citations provides a broad view of influence-a measure that generally correlates with institutional size. The larger the university, the more papers that tend to be produced and, in most cases, the more citations collected. Therefore, it is hardly surprising that the titanic Toronto appears in 18 of the 20 rankings in the total-citations table (landing first in 14 fields). Size also favored the University of British Columbia, which registered among the top three in 17 fields in the total-citations table. On the other hand, assessing citation impact in a given field allows large and small institutions to compete on an equal basis. Larger size does not automatically confer advantage. Thus, in the ranking for physics, for example, smaller institutions such as Simon Fraser University in Burnaby, British Columbia, and Carleton University in Ottawa, Ontario, posted impact scores that placed them above universities that typically overshadow them in other fields.

Similarly, Hamilton, Ontario-based McMaster University's performance in clinical medicine shows how comparatively small institutions can shine when their performance is seen through the lens of citation impact.

In fact, although Simon Fraser registered in only three fields-physics, engineering, and economics and business-it finished first in impact in all three. Similarly, York University in Toronto made the top three in four fields, finishing first in chemistry and geosciences, and second in mathematics and plant and animal sciences. Yet another striking performer was the University of Laval in Quebec City, which posted impact scores high enough for third place in two fields-molecular biology and neuroscience.

Sometimes, however, despite an unexpected performance from an underdog institution, the race really does go to the biggest. The University of Toronto, for example, managed to finish first in impact as well as total citations in seven fields, including materials science, computer science, and biology and biochemistry. And McGill University in Montreal, which ranked in 11 fields in the total-citations table, also scored highly in impact in four fields, including microbiology, clinical medicine, and molecular biology.

Output Thresholds

To make sure that impact was balanced against a reasonable output of papers over the five-year period, Science Watch set a minimum threshold of published papers in each field. This means that some universities, despite high-impact scores, were excluded from the rankings because they published only a handful of papers. In physics, for example, the University of Winnipeg posted an impact score of 5.15, placing it ahead of the University of British Columbia. However, Winnipeg produced just 52 physics papers during the five-year span-well below the threshold of 250 papers set for this field. A similar fate befell McMaster University in astrophysics.

In computer science, Queen's University in Kingston, Ontario, and the University of Ottawa scored well, but fell just short of the threshold for inclusion.

Of course, the volume of published papers varies from field to field. Science Watch attempted to compensate for this in setting an appropriate output threshold for each ranking. This survey examined performance at Canadian universities over the last five years-a relatively short span. To assess impact over a slightly longer haul, Science Watch also tracked performance over the last 14 years in all fields combined (not shown in tables). The University of Toronto showed its prowess by this measure as well, with an average of 11.42 citations per paper. McGill University was next with a mark of an average of 10.45 citations per paper, followed by McMaster University (10.29); the University of British Columbia (9.73); Queen's University (9.07); Dalhousie University in Halifax, Nova Scotia (8.93); the University of Western Ontario (8.82); the University of Montreal (8.50); the University of Sherbrooke, Quebec (8.45); and the University of Lethbridge, Alberta (8.28).

Sidebars

(The Scientist, Vol:10, #2, pg.13 , January 22, 1996)