War of attrition – Asking why PhD students leave

After finishing up a bibliography of sources on graduate education, I wanted to write a post about some of the things I’d read on the topic. Because there had been recent articles about attrition and supervision, in this post I point out the link between them, citing some of the literature on PhD non-completion and its relationship to factors like academic and social integration, professionalization opportunities, and support/mentorship from faculty members. The original post is from July 17, 2013: War of Attrition.

The Times Higher Ed in the UK had a hit this past week, regarding the issue of doctoral supervision, with an article by Tara Brabazon titled “10 truths a PhD supervisor will never tell you”. Worth noting alongside that one is a recent article by Leonard Cassuto that appeared in the USA’s Chronicle of Higher Education, regarding doctoral attrition, which has long been notoriously high (at least in the United States – an average of around 40-50 percent). Attrition rates in Canada are, as far as I know, not generally available though some numbers from eight of the “U-15” were published in this article from Margin Notes blog (and a longer discussion of completion rates and times to completion is here).

I mention these two issues together because for my dissertation I’ve been going over the research on PhD supervision and attrition, including the work of Barbara Lovitts (who’s cited by Cassuto as well), Chris Golde, and Susan Gardner among others. This research shows clear connections between supervision styles, departmental “climates”, professionalization opportunities, “student satisfaction”, and the outcomes of PhD study – including attrition.

What necessitates this research is that there are long-held misconceptions about the causes of non-completion. A key finding is that often faculty attributions of student non-completion have looked very different from either the students’ understanding of their experiences (or of what other students experience), or from the reality of their reasons for leaving. Since those who leave don’t generally get to tell their stories, assumptions can be made that they simply “didn’t have what it takes” or that the admissions committee didn’t “select” the right candidates for the program. Not only does this download the blame onto the individuals who leave, but it also masks other entrenched problems that can then continue without serious examination. Additionally, it doesn’t mesh with research that’s shown the non-completers tend to look just as “prepared” for academic work as the students who finish.

While there is no single reason why students tend to leave (in fact it’s usually a combination of reasons), a major take-away from the scholarship on this topic is that the supervisory relationship is of crucial importance – not only in whether students graduate, but also in their subsequent (academic) careers. For example, Lovitt’s book Leaving the Ivory Tower confirms that supervisors who have already helped PhD students to complete are the ones most likely to continue doing so. However, the reasons are complex. These supervisors tended to have a give-and-take relationship with students rather than expecting the students to do everything on their own. They “scaffolded” and supported their supervisees, and cared about students’ intellectual development and overall well-being; they facilitated the students’ professionalization and their academic and social “integration” into the department and the discipline, through a variety of practices.

If there are no exit interviews with those who leave their programs, then it’s much easier to continue making erroneous assumptions about why they left in the first place. This is important because there are significant policy implications for the reasons we assign for attrition. For example, even Cassuto’s article places emphasis on selection of the “right” types of students, and on certain types of student responsibility such as seeking out the department’s attrition rate before applying – though this is not information that programs tend to provide to potential students. His taxonomy of students doesn’t include those who simply don’t know what support they will need, and don’t end up receiving it; it doesn’t include those who had the capacity to complete but were abandoned by their supervisors, sabotaged by departmental politics, or derailed by personal life circumstances. All these factors are discussed in the literature on PhD attrition.

Like most other issues in education there are many causes for problems with completion. Any relationship is a two-way street, as pointed out in this post by Raul Pacheco-Vega. There are plenty of faculty who are already engaging in the helpful practices described by Lovitts and other researchers, as well as PhD students who don’t put in enough work, or who probably shouldn’t have chosen to start the degree in the first place. But when it comes to implementing solutions, the nature of students’ supervisory relationships should be one of the primary targets of inquiry and intervention.

An example of an important issue that could be addressed is that of responsibility. Reasonable student expectations of faculty should be made clearer, and tacit institutional and professional knowledge – which is so crucial to students’ success in graduate programs – must be made explicit rather than being left to students to discover for themselves. If students understand what they should expect from a good supervisor – and for what they are responsible themselves – they may be able to make a more informed decision about this important working relationship (and whether in fact it’s working at all).

In some cases, this kind of change will take time and a great deal of consideration because if we take the research seriously, the problems extend beyond merely asking professors and students to engage more often in certain practices. They may be problems with the culture of a department or program, or in fact (considering some of the comments from Lovitts’ interviewees) the nature of academe itself, which is where we have to ask ourselves – what kind of a university do we want, and what kind of faculty will be working there? For example, if students also listed “personal problems” (as many of them did) including stress on existing relationships and the demands of raising children, does this mean those who desire a more balanced life will be inherently unsuited to academic work?

PhD students’ “dissatisfaction” should not be dismissed as merely the whining complaints of the academically inadequate. When students don’t know what to expect, they don’t have the opportunity to align their decisions and behaviour with the appropriate expectations; when they don’t receive adequate support, they may not know how to get what they’re missing, or indeed that they’re missing something in the first place (until it’s too late). Not only that but if we ignore these issues, do we not face a reproduction of what may be the worst aspects of academic life, in the name of “trial by fire”? Those who “make it through” are often assumed to have some inherent set of qualities that make them a better “fit” for academic life. But closer attention shows that this clearly isn’t the case, which means – even if the attrition rates are lower in Canada – we need to seek out appropriate explanations.