Building a Database of Institutions

Defining the target group

As 2017 was the first year for the survey, the focus was exclusively on provincially funded and accredited post-secondary educational institutions, which still represent by far the majority of post-secondary institutions and students in Canada.

In subsequent years the intent is to widen the focus, to include post-secondary institutions that are federally funded, institutions managed by First Nations, and private colleges that receive significant public funding.

Sources

One challenge the survey faced was the lack of a commonly used, publicly accessible database of all Canadian public post-secondary educational institutions. We worked our way through the membership listings of Universities Canada, Colleges and Institutes Canada (CICAN), Maclean’s EduHub, and provincial government web sites. From Statistics Canada we could find only aggregate data on student enrolments broken down by province and by part-time or full time students, but not data for individual institutions.

Selected institutions

We ended up with a list of 203 institutions, once we had eliminated duplications, incorporated affiliated colleges and universities with the main institution awarding the qualification, and removed institutions not funded by provincial governments. You can find a link to the list of the 203 institutions that formed the population base for the 2017 survey here.

This was not an easy task and required some judgements on our part. Some of the universities for instance have colleges that over time have been merged with or affiliated to the main university. Some universities were bi-lingual, and some Anglophone universities had a francophone college. Other institutions have devolved from an older university into a semi-autonomous institute, although the degree transcript still comes from the main university. Should we count each of these as separate universities or at least send them separate questionnaires? This was particularly difficult for Québec universities.

In the end, we decided that it would be sensible to send one questionnaire to the main university, and ask them to complete it as best as possible to include all their affiliated colleges and institutes. We then had to identify the appropriate contact person in each institution (usually Provosts or VPs Education).

Key variables

We also identified institutions by language (anglophone or francophone) and their total student headcount (full-time and part-time), almost entirely from information publicly available through provincial government web sites, although not all provinces provide this information. Where this was the case, we were usually able to find an institutional annual report or government audit that provided ‘official’ enrolment figures for the institution.

Result

For the 2017 survey, this process resulted in

• 72 universities (35%),
• 81 colleges outside Québec (40%), and
• 50 CEGEPs/colleges within Québec (25%).

Of the 203 institutions, 70 (34%) were either francophone institutions or were bi-lingual institutions with a separate francophone program.

List of institutions in the 2017 population base: note: not all of these institutions participated in the survey –see Response rate in the Public Report, pp.58-60.

Complications

One thing that became clear even at this stage is that there is no consistency between provinces, or between provinces and Statistics Canada, on how information about students is collected or reported. Several different measures are used: student headcount (full time, OR full time and part-time); student course enrolments; student FTEs (full-time equivalents); and student program enrolments, with variations within each of these broad categories. Also some data include non-credit, continuing education students as well as students taking courses for credit.

This variation in student statistics makes inter-provincial comparisons very difficult. In the end, for the database of all institutions for the survey, the population base, we used primarily official provincial student headcounts for individual institutions for students taking credit courses or programs, the measure most common across all provinces.

Nevertheless, there were still differences between provinces in the latest year for which data were collected, whether the data included part-time as well as full-time students or out-of-province students, and in some cases we believe non-credit students were also included, but it was often impossible to tell.

Establishing an independent comparator

To validate the representativeness of our survey respondents, we looked for independent data to which we could compare our database and survey responses.

For the 2017 survey, Statistics Canada’s most recent figures for Canadian post-secondary student enrolments were for the fall of the 2014/2015 academic year. Statistics Canada’s enrolment numbers were based on program counts and not student counts. If a student is enrolled in more than one program as of the snapshot date, then all of their programs are included in the count. Nevertheless, this was the closest independent comparator we could find.

Comparing our population base with StatCan, we found the following differences in student enrolments.

Without knowing more about the basis on which Statistics Canada built its data, we cannot explain the difference between the two populations sets, but the differences are relatively small, except for CEGEPs.

We are confident we have included all the CEGEP institutions but we probably do not have all enrolled students counted, just those for which the Québec provincial government provides funding, from which we derived the data. Nevertheless, if we take Statistics Canada data as the comparator, our population base appears to represent a very large proportion (93%) of students studying for institutional credit at Canadian public post-secondary institutions.

The minor changes were are making to the population base for the 2018 survey should make the survey even more representative.