Cosupervision: Risks and Approaches
Term is over, and I have the time to share some of the scholarly work I’ve been doing with the OxICFM Centre for Doctoral Training. I prepared a short presentation to supervisors about cosupervision, and hope it may be of use to people beyond Oxford.
It’s an odd thing, telling supervisors about supervision. I have never supervised a doctoral student, so I clearly know much less about the practicalities of supervision than anyone who actually does it. At the same time, most academics have no chances to read the educational literature.
This means that I have something valuable to offer supervisors: not a prescriptive ‘how to’, but a descriptive ‘how might’. I am not trying to tell anyone how to supervise, but my experience is that all supervisors are keen to supervise well; I hope that something in here sparks some ideas.
Cosupervision is becoming quite common in the UK. Some of this is led by the nature of specific projects: if the object of investigation is interdisciplinary, then sharing a student between supervisors with complementary expertise seems very sensible. Some of cosupervision’s popularity is more economic: it is a blunt requirement of many funders, and cosupervision can mitigate institutional risks (e.g. when supervisors have become more mobile as a result of structures such as the REF). There is also tremendous scope for a cosupervision model to broaden the perspective of the student, helping them to see different ways of being and to develop sophisticated interpersonal skills.
But if cosupervision is becoming more common, it remains very difficult. Cosupervision is a significantly different activity to supervision, and succeeding as a cosupervisor is not straightforwardly related to succeeding as a supervisor.
Pole has a background in investigating the parameters of single-supervisor supervision quality in UK science doctorates, and determined 9 themes which were important for the quality of supervision. The paper notes that the sample composition did not let the results speak to whether ethnicity mattered. (In my opinion, it must.)
Broadly, Pole showed that everything matters. The relationship between a student and a supervisor relates to technical details relating to the research and doctoral processes, but also to the identities of both parties. This is complicated, and though the parameters of cosupervision quality seem likely to parallel those for single supervisors, adding an extra person into the mix certainly won’t simplify things.
Akerlind and McAlpine looked at the ways that different supervisors thought about what a doctorate is. Though this was not in a UK context, their findings ring very true. Supervisors adopted one or two of the three views about what a doctorate is.
Does a doctorate produce independent researchers? In that case, the appropriate supervision style is to help students develop their decision making skills. This may include supporting students to make the choices that you – an experienced researcher with an intuition about what will and won’t work – might not have made.
Does a doctorate foster imagination? In that case, the right way to supervise might be to encourage people to think deeply about the field. Reading and discussing papers might be the core focus of a supervisor’s support to a student.
But what is the purpose of a dcotorate is to develop the student as an individual? The idea of developing an individual was slightly more fuzzy, and related to professional career trajectories, and might relate to the CV objects which support doctoral students to become postdocs (predominantly papers). Supervisors with this conception of a doctorate might focus on motivating students so that their published output is competitive.
It is possible to adopt more than one of these positions, but it is clear that trade-offs must be made between them: a student can’t run columns while they’re reading papers.
More profoundly, though, it is very possible that different cosupervisors might tacitly adopt different positions. It’s often hard to articulate these broad, thematic conceptions of a doctorate; this risks inadvertently pulling a cosupervised student in really different directions.
In this light, it is interesting to think about how students experience doctorates. A recent HEPI report notes that the PG student experience needs to be investigated much more rigorously, but a few objects are useful for considering cosupervision.
Most students think about a doctorate as a big masters project at first. Initially, they do not understand the responsibility and initiative required in a doctorate. Their expectations solidify during the first year, when they get a ‘feel’ for how research groups operate. Clearly, both of these ideas will be influenced by a cosupervision setting.
Three specific interventions from the literature might be worth considering when cosupervising. Discussing the roles of all parties (and the mutual expectations) can help scaffold the way that the student comes to understand cosupervision. Focusing these discussions onto the specific dimensions of the doctoral project can help to make this conversation more concrete than philosophical. Finally, reviewing the cosupervision relationship regularly near the start of the project can be useful.
I have tried to embody a few of the specific ways that students can be put at risk by the cosupervision model. These ideas are my reading of the broader literature; I present them as imaginative stimulus rather than hard fact.
DIFFERENT WORLDVIEWS
If one supervisor thinks that doctorates are about getting crystal structures of every compound and the other thinks that it’s about really understanding the literature, the student is presented with thematic expectations which may be unreconcilable: there is not ‘big picture’ of what a doctorate is, because there are two ‘small pictures’.
DIFFERENT AGENDAS
Closely related is the idea of agendas. If one supervisor is aiming to develop a really deep understanding of some system while the other is pursuing a more incremental programme (e.g. producing papers or patents) then the patterns of work for these agendas may come into conflict with each other.
POWER DYNAMICS
This can mean that students are the site of supervisor-supervisor conflict. Students are valuable resources, and competition for the focus of the student can be uncomfortable for the student themself. Avoiding this by coming to some supervisor-supervisor arrangement can be experienced as a ‘piggy in the middle’ situation, where the student is excluded from the process of making decisinos.
SEEKING CONSENSUS
Faced with the demands of two supervisors, many students simply try to fulfil everyone’s expectations. This can have the effect of students pursuing the experiments which everyone agrees on. This mechanism can discourage taking appropriate risks in research, leading to a safer doctorate.
THE TRANSLATOR
A final risk is that students in interdisciplinary settings may become a neutral ‘translator’ between two supervisors, rather than an independent actor in their own right.
In my presentations to staff, The Translator provoked the most push-back. Several colleagues described how they derived tremendous value from being a translator during their own doctorates. I think this emphasises a broader point with all of these risks: each of them is the shadow of an opportunity. Aspects of all of these dangers could be realised in a really positive way in cosupervision.
Cosupervision is growing in popularity, and that means that staff are likely to be faced with increasing cosupervision responsibility. We have an interesting chance to grow this as part of our culture rather than to have it imposed upon us as some strategic decision.
I have seen strategically-imposed cosupervision fail catastrophically, perhaps because it places the student in such a vulnerable position. At the same time, I was myself very successfully cosupervised (thanks, supervisors!). My dearest hope is that this presentation leaves you with a sense that cosupervision is really tricky, and that discussing it with colleagues is something we could usefully do from time to time. This focus on a developing a shared culture of successful cosupervision is the way we can best support our students.
One specific strategic object can be extremely useful in supporting students, though: the design of the project. It is hard to say exactly how this will apply to your fields of research, but it seems likely that there are certain shapes of project will result in projects with a higher chance of successful cosupervision. Defining the roles of the supervisors is also an extremely high-value activity.
The typical cosupervision model in the OxICFM CDT seems like one sensible design. Students are mostly in one group, and then work with cosupervisors for well-defined sub-projects (e.g. specialist characterisation or catalytic studies of a synthesised compound). This structure allows students an authentic experience of cosupervision, but also leverages the strucures of support implicit within research groups. In my view, this could work very well if supervisors communicate well with each other and the student.
In summary, the best way to cosupervise well is to talk. The literature suggests that the main problems arise from poor supervisor-supervisor communication, and that the most effective single thing you can do when cosupervising is to have regular conversations between all parties.
A case in point has been the way that supervisors have raised the role of postdocs in supervising doctoral students when discussing their experiences of cosupervision. This is something I hadn’t thought about, but is clearly an extremely important part of the culture of research groups in academic Chemistry.
