University of Toronto

 > U of T Home > About U of T > Measuring Our Performance > University Rankings and Faculty Data

University Rankings and Faculty Data

July 18, 2008

Introduction

The University made public its position on data-gathering for institutional ‘ranking’ and performance measures in 2006. As the national leader in performance measurement we indicated our intention to continue developing and posting on-line more information for others to analyze as they see fit.

The University has already played a catalytic role in the creation of CUDO (Common University Data - Ontario). Multiple universities are now collaborating nationally with a view to developing a parallel national data set with common definitions and measures.

Common definitions are particularly important, because a lack of clear and shared definitions for variables leads to both random and systematic errors in simple and composite measures. These errors occur even in what, on the surface, should be straightforward items, such as counts of the numbers of faculty and students in a given institution.

Faculty Counts and Student-Faculty Ratios

The student-faculty ratio is calculated crudely as: student headcount / faculty headcount.

The ratio can be adjusted for full-time equivalents as follows: FTE student count/ FTE faculty count.

Through the years, universities, governments, performance-rating agencies, and media outlets have all reported these ratios. The ratio is sometimes portrayed as a proxy for ‘educational quality’, on the grounds that a lower ratio connotes more opportunities for students to interact personally and directly with their teachers. It is also sometimes presented as a measure of how institutional resources are being deployed.

Unfortunately, the interpretation of student-faculty ratios is anything but straightforward.

First, student-faculty ratios are not direct measures of educational quality. At best these ratios are input measures. They may not correlate closely with student experiences or learning outcomes. Other measures such as student survey results and graduation rates provide more meaningful information to assess the quality of the learning environment.

Second, as noted, the ratio depends on both student and faculty counts. The impact of errors in these counts is unpredictable. Sometimes random errors will cancel out. At other times, the ratio may be substantially skewed.

Third, counting students should be relatively straightforward. However, we note that student counts rise if an institution includes learners who are not registered in the usual ways (e.g. individuals pursuing diplomas and certificates, those taking continuing education courses, or postgraduate medical trainees). Proper adjustment for full-time equivalents is also appropriate.

Fourth, tallying faculty is surprisingly challenging, because there are many different categories of academic appointees and many ways to count them. The range of categories is greatest for institutions with professional schools or affiliated research institutes. Faculty can be categorized by appointment status (e.g. tenure-stream, teaching-stream, short-term contract, adjunct), by rank (e.g. assistant, associate and full professors), by time commitment (full-time, part-time), by job description (e.g. research scientists, clinical faculty), or by salary source (university or affiliated institution). What these categories mean in terms of contribution to the teaching and research mission of the University also varies from one institution to the next.

Fifth, even if there is an agreed definition for overall faculty counts, there are still problems estimating full-time equivalents. Research scientist ‘X’ may hold a faculty position but contribute only to student supervision in a laboratory setting. Adjunct professor ‘Y’ may teach a popular half-course. And short-term instructor ‘Z’ may be teaching two full courses a year, equivalent to a full-time tenured professor. Estimates of full-time equivalents, while difficult, are nonetheless important for comparative purposes.

Sixth, at least in Canada, the sources of data currently available to the public have severe limitations. In the past the University of Toronto has relied upon the Statistics Canada faculty survey and its classifications in presenting our faculty counts. However, these counts were developed in large part to facilitate collection of salary data, and fail to reflect the broad array of faculty who contribute in a meaningful way to our institution.

In responding to external agencies that seek data on faculty counts to calculate student-faculty ratios, some Canadian institutions have continued to use the Statistics Canada tally of instructional faculty on the full-time university payroll, or some close approximation to that count. In cases where definitions are open to interpretation, institutions have taken different approaches. For example, Tables 1 and 2 below show variability in the 2007 data submitted to the Times High Education Supplement by North American and UK institutions. Student-faculty ratios therefore varied widely across institutions. The tables also show marked intra-institutional variability resulting from different counting methods. Unfortunately, these same ratios were incorporated into scoring systems and used to generate ‘rankings’, underscoring both the conceptual problems of composite scores and a specific need for reliable data definitions.

Table 1: Student-Faculty Ratios: Canada & US
	S-F Ratio Using Publicly-Available Data	S-F Ratios based on data submitted to THES			2007 THES S-F Ratio Score
		Based on Student HC / Faculty HC	Based on Student FTE / Faculty HC	Based on Student FTE / Faculty FTE
A. Canadian Institutions
University of Toronto	23.7	24.5	21.1	22.3	21
University of Alberta	21.3	23.5	20.8	20.8	23
University of British Columbia	16.3	10.6	8.4	Not Provided	70
McGill University	17.0	5.3	Not Provided	Not Provided	99
McMaster University	18.1	18.2	Not Provided	Not Provided	31
Queen’s University	22.5	12.6	11.2	Not Provided	49
University of Western Ontario	22.2	18.6	Not Provided	Not Provided	30
University of Montreal	23.4	18.2	16.8	16.8	31
B. US Institutions
University of Washington	11.2	9.4	7.8	Not Provided	73
University of Pittsburgh	17.3	7.1	6.2	Not Provided	94
Ohio State University	12.0	16.4	14.8	14.9	39
University of California-Berkley	15.1	12.3	11.8	Not Provided	59

[1] All Canadian Institution S-F Ratio are developed using 2004-05 Statistics Canada FT Faculty and FTE students data. All US Institutions (except UC Berkley’s) S-F Ratios are as per the 2006-07 Common Data Set. UC Berkley’s data is 2007-08.

The US Common Data Set Definition states: FTE students to FTE instructional faculty excluding student and faculty in stand-alone graduate of professional programs such as medicine, law veterinary, dentistry, social work, business or public health in which faculty teach virtually only graduate-level students. Teaching Assistants excluded.

Table 2: Student-Faculty Ratios: UK Institutions
	Publicly-Available S-F Ratio*		2007 THES
	FT Teaching & Research Faculty*	FT Teaching & Research Faculty plus Research only Faculty*	2007 S-F Ratio Based on FTES	2007 THES S-F Ratio Score
University of Cambridge	10.9	4.2	4.9	99
University of Oxford	11.7	4.3	4.7	100
Imperial College of London	10.2	9.2	4.1	100
University College London	8.7	3.9	4.4	100
University of Edinburgh	12.5	6.5	8.0	82
King’s College London	10.8	5.4	6.8	91
University of Manchester	12.7	7.2	8.5	77

*Source: 2000-01 HESA Data

The University of Toronto has therefore been working towards a more comprehensive and appropriate summary of the various faculty members who contribute to our academic activities.

As published several months ago in our Performance Indicators Annual Report and shown again in the figure below, our faculty counts vary dramatically depending on definition. In the past, we have focused on reporting to the public our counts of core instructional faculty. For our institution that includes full-time tenure or tenure-stream faculty and teaching-stream faculty. But, as indicated below there a thousands of other faculty that contribute to the teaching and research mission of the university. Using updated 2007 data, we note that this variability is reduced as one adjusts for full-time equivalency (see Table 3). Nonetheless, our effective student-faculty ratio which runs well above 20 with a restrictive definition, falls to 10 with a complete count and adjustment for full-time equivalents.

We are encouraged by the progress to date in refining our total counts and full-time equivalents. Additional work is needed, of course, to delineate the contributions made by these different categories of faculty members to the educational and research missions of the University. It will be important for national data consortia to develop consistent faculty definitions and FTE adjustments to enable more informative comparisons of both the teaching resources available to students and the research capacity of institutions.

Table 3: Total University of Toronto Faculty, Fall 2007
Faculty Category	Full-time	Part-time	Total	Total FTEs(e)
Professorial (Tenure/Tenure-Stream) Faculty *	2,093	166	2,259	2,183
Teaching-Stream Faculty	293	117	410	359
Clinical Faculty	1,971	1,703	3,674	2,216
Term-limited Instructional Faculty (Sessional, Stipendary)		924	924	185
Research Appointees (Research Scientists, Other Graduate Supervisors)	247	544	791	364
Non-Payroll Faculty (Status-only)	344	1,597	1,941	663
TOTAL			9,999	5,970

*Includes Non-Tenure Stream Faculty.

Faculty Counts for Ranking Publications

As noted above, omnibus ranking exercises are characterized by attempts to aggregate many aspects of the work of complex institutions into a single composite measure. There is no strong basis in measurement theory for attempts to distill and combine research, educational and other elements of university performance into one numerical score or ranking. On the other hand, the information-gathering exercises by publications or agencies that undertake ‘rankings’ have offered valuable opportunities to compare individual lines of existing data across institutions. In order to improve data availability and comparability, we are providing faculty counts, based on the above analysis, in response to our understanding of different survey requests.

THES-QS

"Total number of academic faculty staff who are responsible for planning, directing and undertaking teaching only, research only or both teaching and research. Please include: vice-chancellors, deputy vice-chancellors, principals, professors, heads of school, associate professors, principal lecturers and tutors. Please exclude research assistants*, PhD students who contribute to teaching, and exchange scholars or visiting professors who are members of another university."

Total 2007 Faculty Count: 9,999

Total 2007 Estimated Faculty FTEs: 5,970.

Maclean’s

“The total number of full-time instructional faculty members in each of the following categories for Fall 2007.”

Total number of instructional faculty*: FT 2,644; PT (FTE) 160.4

Total number with doctorate, first professional, or other highest level degree: FT 2,418.

*Note: As per Common University Data Ontario (CUDO), includes Professorial Rank in the Tenure/Tenure-stream and Non-Tenure-stream and Teaching-stream faculty on contracts of 12-months or more. Excludes those on contracts of less than 12-months, non-appointed stipendary instructional staff and clinicians not paid directly by the University.