Introduction

View the most recent version.

When students begin their postsecondary education (PSE), their subsequent schooling can take myriad pathways. Some students continue in the same program until graduation. Others switch to a different program at the same institution. Some change institutions, either at the same level of study (college, university, private training), or at a different one. Still other students abandon their studies, some altogether and others to return at a later date. Some finish one program and stop there, while others continue on with their studies, either directly or after a break.

Understanding such pathways is important for a variety of reasons to university and college administrators, to Kindergarten to Grade 12 (K-12) and postsecondary education policy makers, to academics, and to others, including parents and students themselves. Our understanding of these processes is quite limited, however, largely due to the general shortage of data suitable for studying dynamics that can be both detailed and complex.

The general objective of this paper is to provide new and unique empirical evidence on PSE pathways in Atlantic Canada based on the Postsecondary Student Information System (PSIS), which has been developed at Statistics Canada from student administrative files provided by postsecondary institutions. Atlantic Canada is especially well positioned to take advantage of the PSIS – for this study as well as for others that could be undertaken on this same topic or others – due to their strong participation rate in PSIS since 2001. The longitudinal linking of student records that has been carried out by Statistics Canada for this project broadens the possibilities for further research (see section 4.1 for a discussion of the longitudinal linked file).

The PSIS data used in this project covers four reporting years: 2001/2002 through 2004/2005 (more years are scheduled to be added on an on-going basis).¹ Most importantly for the purposes of this project, the PSIS data include one record for every student for every PSE program in which they were enrolled in a public PSE institution in Atlantic Canada for each of the four years covered by the data. This is possible only because of the almost complete coverage of the PSIS throughout Atlantic Canada regardless of the level (college or university), institution, or province of study.

This information has allowed for a longitudinal tracking of students which allows us to follow individuals from their point of entering PSE through the ensuing PSE dynamics of primary interest, which we have categorised as i) graduation, ii) continuing their studies at the original institution, iii) switching to another institution, possibly at a different level (i.e., college versus university) or even in another province (as long as it is in Atlantic Canada, which is the geographical limit of the longitudinal linked file), and iv) leaving PSE without graduating.

We also identify the proportion of leavers who return to PSE after leaving, again whether at the original institution or at another, at the same or at a different level of study, in the first province of study or in another within Atlantic Canada.

Finally, the analysis also includes a tracking of those observed to graduate from a program over any of these fours years to see how many continue in their studies – either directly or after taking a short break – and if so, where they do so (level, province of study, etc.).

The data utilized for this analysis covers PSE students attending public institutions at all levels of study: college, bachelor's degrees, master's students, Ph.D. candidates, and those entering first professional degrees. The emphasis is, however, on college and bachelor's students, with the latter three groups (master's, Ph.D., first professional programs) included only in the first part of the analysis (the initial hazard transition rates in the first years of their programs). This focus on college and bachelor's students is due to, first of all, the relatively small sample sizes for the other groups, but also due to the nature of PSE experiences at these levels (e.g., master's, Ph.D. and professional students are much less likely to switch programs or leave PSE once they have started). Finally, students at the college and bachelor's levels are presumably the main focus of interest and concern for administrators, policy makers, and others, particular in terms of the "persistence" dynamics focussed on in this study, precisely because of their higher rates of switching and dropping out.

In covering this full set of PSE dynamics for the whole of Atlantic Canada with a fully representative sample (essentially the entire population of students), this study is the first of its type for the region. It is also, in most respects, unique in Canada, and is notable even at the international level.² The results reported here might therefore be of interest to academics, institutional administrators, postsecondary policy makers, and others with an interest in these dynamics not only in Atlantic Canada, but across Canada, and possibly even in other countries. We hope the paper will, in addition, illuminate the principal strengths – as well as the limitations – of the PSIS data, and thus help identify not only the potential extensions of the current work, but also other potential research projects that could be undertaken with these new and extremely rich data.

The paper is organised as follows. In Section 2, we review the previous research on persistence in PSE with a focus on Canadian work. Section 3 explains the methodological approach. Section 4 describes the data and samples used. In Section 5 the empirical findings are presented. Finally, Section 6 summarises the main findings, places them in a policy context, and mentions a number of possible routes for further research.

---

Notes

1. For each institution, the underlying annual files cover one calendar year of student records, thus running 2001/2002, 2002/2003, 2003/2004, and 2004/2005. The greatest number of institutions' reporting periods run from April 30 or May 1. Others' begin in another day in April or May, or in June, and one starts July 1.
2. The authors are also engaged in research of a similar, complementary type using Statistics Canada's Youth in Transition Survey, their papers including Finnie and Qiu, forthcoming, 2008.