Given the choice, most PhD students would prefer to receive individual supervision rather than be supervised alongside their peers as part of a collective. This is understandable, given the undivided attention and precise, directly relevant advice one would receive. However, Hans Agné and Ulf Mörkenstam have compared the experiences of individually and collectively supervised students on the same doctoral programme and found that collective supervision, during the first year at least, is correlated with significantly shorter times to thesis completion compared to individual supervision.
Imagine you are about to begin your doctoral research, either for the first time or in a new field of research. Would you prefer to be supervised individually, where a single expert meets exclusively with you and offers suggestions motivated uniquely by your doctoral research? Or would you instead prefer to be supervised, at least in the first year, together with other doctoral students who meet jointly and regularly with a group of experienced researchers for advice and feedback on the work of the whole group?
If you are like most of us, in that situation you will prefer individual rather than collective supervision. Why shouldn’t you? It seems natural to think your research will benefit most from learned reactions motivated exclusively by your research. Reactions prompted by the work of other doctoral students may, depending on your personality and role in the group, seem like a waste of valuable time. Moreover, a supervisor or group of supervisors selected to guide your work, and your work only, will likely have greater expertise that is directly relevant to your needs, compared with supervisors chosen to meet the needs of your co-doctoral students as well.
But if this is indeed your preference then I’m afraid you are in trouble. If your department allows you to choose individual rather than collective supervision in your first year, you are likely to be acting against your own interests. Our recent research suggests that individual supervision, at least in the first year, will lead you to lose time to undertake new research or otherwise advance your career. Moreover, by committing to your individual supervision your department may use more time and resources than if you were to be supervised as part of a collective.
Before arriving at this conclusion, we investigated 145 doctoral students in political science at Stockholm University, admitted between 1991 and 2014. When it comes to assisting doctoral students to complete their theses on time or even as quickly as possible, this research has led us to believe that collective supervision in the first year of study significantly out-performs individual supervision. But before detailing our results and procedures, let us rewind and provide some background.
Supervision of independent research projects is a key practice at universities worldwide. It is used to transfer knowledge among individuals and encourage the development of new ideas, as well as for a range of other purposes. That said, there has been little systematic research into which kinds of supervision are more effective than others at attaining any given set of objectives. Research has devoted much time to describing a variety of ideals, problems, and practices in supervision, while not actually testing the effects of the differences observed. Individual and collective supervision is one of several critical but as yet untested distinctions.
Quite frustratingly, but also interestingly, previous research holds competing expectations regarding the consequences of individual and collective supervision. While doctoral students often have their reasons to prefer individual over collective supervision, there are plenty of arguments that point in the opposite direction. The research literature suggests that collective supervision may enhance peer learning, broaden the academic learning context and the pool of knowledge, facilitate acquisition of the values and behaviours of a research practice community, reduce the risk of linking doctoral students with a single supervisor before topic selection has been finalised, and resolve disagreements among senior staff responsible for providing supervision. Such factors would seem to shorten, not prolong, the time to completion.
To examine what actually happens in different kinds of doctoral supervision we found ourselves in the lucky position of having been directors of a doctoral studies programme which, back in 2009, shifted its teaching model from individual to collective supervision during the first year. We could therefore create data which makes it possible to test empirically whether the introduction of collective supervision in this case had a positive, negative, or no effect at all on the time taken for doctoral students to complete their theses. All we had to do was compare the time to completion of doctoral students admitted before and after 2009, and, of course, to control for the alternative explanations we knew from our own experience and existing research are likely to affect time to completion. For example, it is well-known that time to doctoral completion depends on funding opportunities, the academic discipline, and the integration of doctoral students into ongoing research projects.
Studied in this way, it appears that collective supervision of first-year doctoral students is correlated with significantly shorter times to thesis completion compared to individual supervision. Students who received collective supervision in the first year averaged 57 months in the programme, while students receiving individual supervision averaged 92 months. These differences, and some others, are depicted in Figure 1, below. Multivariate statistical analyses of these observations support the conclusion that collective supervision in the first year does indeed reduce the time to completion of doctoral students.
Figure 1: Mean number of months for different groups of students from admittance to defence of doctoral theses. This figure is taken from “Should first-year doctoral students be supervised collectively or individually? Effects on thesis completion and time to completion”, published in Higher Education Research & Development (2018).
We believe these results constitute a strong argument for further collection and analysis of similar data in other contexts – for instance in other countries and in other disciplines – but also that they yield new and relevant support for general assumptions about the effectiveness of collective supervision during early teaching phases, such that it strengthens opportunities for peer-learning among doctoral students. We believe, therefore, that collective supervision, at least in the first year of doctoral studies, is superior to individual supervision, when it comes to reducing time to completion, and potentially also for other, higher aims in doctoral research, (e.g. to create innovative ideas and to analyse them in novel ways).
In light of this conclusion about the effects of alternative supervision models, we would like to end with two caveats on its practical application. Firstly, focusing too closely in policy discussions on time to completion may risk placing an unnecessary burden on doctoral students who are already under stress, adversely affecting both their health and their learning. Secondly, our study does not analyse the scientific contribution of doctoral programmes or theses. We do not know whether such factors are related to time to completion, or in what way (for instance whether fast progress correlates with superficial research). So, in contexts where most doctoral theses are completed within time limits, the right focus of policy discussions may well be to encourage doctoral students to select more risky, creative, and demanding research problems rather than to shorten time to completion. Whether collective supervision is more effective than individual supervision in attaining those higher aims in doctoral studies is another key problem to be addressed by future research.
This blog post is based on the authors’ article, “Should first-year doctoral students be supervised collectively or individually? Effects on thesis completion and time to completion”, published in Higher Education Research & Development (DOI: 10.1080/07294360.2018.1453785).
Featured image credit: Workplace, via Pixabay (licensed under a CC0 1.0 license).
Note: This article gives the views of the authors, and not the position of the LSE Impact Blog, nor of the London School of Economics. Please review our comments policy if you have any concerns on posting a comment below.
About the authors
Hans Agné is associate professor and senior lecturer in Political Science at Stockholm University. He has vast experience of individual and collective supervision as well as teaching at the doctoral level, and was Director of the PhD programme analysed in this article 2013-17. His main research interests include democracy and legitimacy in the fields of political theory, international relations, and the comparative politics of globalising states.
Ulf Mörkenstam is associate professor and senior lecturer in Political Science at Stockholm University. He has vast experience from supervising students on the doctoral level, as well as on the master and graduate level, and was Director of the PhD programme 2000-09. His main fields of research are minority rights and the history of ideas. The last years he has been principal investigator for several projects on the rights of indigenous peoples with a specific focus on the Sámi people’s right to self-determination.
Really fascinating study. Hopefully the camaraderie that may come from collective supervision can have positive impact on wellbeing and combating feelings of loneliness felt by many doctoral candidates.
What exactly is meant by “collective PhD supervision”?
A very interesting article. Collective supervision may create a discourse among candidates and create a platform to share on methods and methodologies of research. At times, individual candidates may be frustrated figuring out the direction of their research. From my experience, I benefited a lot by discussing my own research with other similar candidates – they usually gave me more encouragement to continue, and also provide some ideas and opinions in how to tackle some research hiccups.
I think in the experimental sciences doctoral candidates work in a lab which often provides (informal) group supervision.
It is routine in North American universities for PhD candidates to have a dissertation advisor who also convenes meetings of their superisees every week or fortnight.