Should I stay or should I go now? The postdoc mobility myth


First published at physicsfocus.

Far back in the mists of time, in the great and glorious days of Britpop, Forrest Gump, and John Major’s “Back to Basics”, I was a postdoc[1]. It was my first postdoctoral position after completing my PhD at Dublin City University in late 1993 and I loved the research I was doing, the working environment, and the camaraderie and teamwork of the Nottingham group.

But I hated the underlying volatility of the post.

After completing a two-year postdoc, I was funded by consecutive short-term contracts. At any time the funding stream could have dried up, and I would have had to move on. I got lucky: I secured a permanent lectureship post after three years at Nottingham and I’ve stayed there ever since – the School of Physics and Astronomy is a fantastic place to work.

I know for a fact, however, that the research ‘outputs’ I had in 1997 – enough for a lectureship at the time – wouldn’t get me within sniffing distance of a shortlist today. The bar has been raised dramatically for postdocs over the intervening years. Increasingly, the route to a permanent academic position involves first winning a fellowship through a highly competitive process.

One of the factors which is very often taken into consideration when selecting for both fellowship and lectureship positions is the “mobility” of the candidate. Indeed, the Leverhulme Trust now explicitly states in its advice to applicants that mobility is a key criterion: “Priority will be given to applicants who show evidence of mobility during their academic careers to date.”

The blunt statement that mobility will be used as a criterion in selecting fellows – with no attempt to qualify this in terms of the personal circumstances of the applicant – reveals some worryingly simplistic and out-dated thinking from the Trust. They are not alone, of course, in assuming that mobility must necessarily be an advantage for a researcher, as this recent article points out in the context of EU funding programmes. The arguments about mobility in that piece resonated with me because I coordinate a Marie Curie Initial Training Network (ITN) project which funds 14 early-career researchers across six countries. Researcher mobility for ITN projects is not only advantageous from the point of view of the European Commission, it’s essential – I can’t employ a UK national on a Marie Curie ITN contract in Nottingham. (Can someone please make sure that this nugget of EU funding policy wings its way to Nigel Farage? I want to watch him spontaneously self-combust…)

The argument that is often made – and which was voiced during a lengthy twitter debate with my Head of School[2] and others on this topic yesterday – is that a postdoc, let’s say Dr. Globetrotter, who has moved from group to group is likely to have greater drive, motivation, and scientific independence than her colleague, Dr. Stayen-Putt, who has remained at the same institution throughout her undergrad, postgrad, and postdoctoral career.

I really don’t buy this argument at all.

Skewing the selection process towards candidates who are willing to ‘up sticks’ and move to a new group every few years immediately disadvantages – and, at worst, discriminates against – those whose personal circumstances and family commitments mean that they do not have the freedom to move. I, for one, would not have been willing to disrupt my children’s lives on a regular basis simply so I could demonstrate a commitment to mobility to a fellowship panel. And I find it rather insulting that this could have been interpreted as a lack of scientific drive, motivation, and independence.

The assumption that scientific independence correlates positively with mobility also needs to be challenged. There is no evidence at all that a postdoc who has been in the same institution for their entire career is any less scientifically independent, or any less scientifically motivated, than a researcher who clocks up the air miles. Indeed, I can think of reasons why there could be a negative correlation between mobility and scientific independence – it takes considerable time to establish oneself at a new institution, to learn to interact with a new group of colleagues, and to work out how you can carve out a niche to “make a mark”.

Moreover, there’s a rather straightforward, pragmatic reason why mobility may not be conducive to establishing scientific independence. Experimental physics is not easy – the ‘kit’ is often complicated and frustratingly temperamental (particularly for non-commercial systems which the researcher has built themselves). If the experimental infrastructure in an institution is very well-matched to a researcher’s scientific goals it would be perverse for them to move simply so that they can tick the mobility box.

And finally, the wonders of the interwebs mean that researchers are connected like never before. In this context, the Leverhulme Trust’s focus on mobility as a criterion in awarding fellowships is particularly quaint, given the extent to which research groups now network and interact virtually.

Image: Britpop, an automatic association with the early 90s – but the bar has been raised for postdocs since then. Credit: Danny PiG/Flickr. Reproduced under a Creative Commons licence.

[1] …and a Douglas Adams fan.

[2] My P45 is in the post.

Confused? Good. You might be about to learn something.


First published at physicsfocus.

It’s never the most comfortable of feelings to have some aspect of your work described as “appalling”. But that’s what greeted me yesterday morning when I scanned down the comments thread for the most recent video I’ve made with Brady Haran for his Sixty Symbols channel


The source of the opprobrium? Well, the video in question was on the topic of entropy. I should have learned by now not to go near the topic of entropy with a barge-pole for a YouTube video because it’s a subject that really can’t be done justice in five or ten minutes. But I had attended a brilliant and inspiring colloquium by Daan Frankel on entropy and self-assembly shortly before the video was filmed. As part of his talk Daan had described the pioneering work done by Sharon Glotzer’s group at the University of Michigan on the role of entropy in the self-organisation of nanoparticles. Glotzer’s group has neatly shown how entropy can be exploited to drive an ensemble of nanoparticles to an ordered state.

Yep, that’s right. Entropy produced order, not disorder. (I enthusiastically recommend Glotzer’s TEDx talk for more on this.)

I thought that this departure from the traditional view of the role of entropy would make a great subject for a Sixty Symbols video and suggested it as a topic to Brady. We filmed it and, as ever, Brady and I had some healthy and robust debate about the role of analogies and metaphor in explaining the physics. Overall I was pretty happy with how the filming went. (As I’ve discussed elsewhere, we academics do not get involved with the editing of the Sixty Symbols videos – that’s all expertly done by Brady. The first time we see the finished product is when it goes online.)

So why does YouTube commenter TheRumpus feel so strongly that the video doesn’t work? Well, you can of course read his comment for yourself but it was the final two lines which particularly resonated with me (for reasons I’ll go into below):


As I’ve explained over at YouTube, Sixty Symbols videos – certainly those with which I am involved – are not meant to be tutorials or mini-lectures. No-one should expect to come away with a solid understanding of entropy on the basis of watching a YouTube video (otherwise why would we bother with setting problems, coursework, lab work, and/or projects for thermodynamics courses – or, indeed, any aspect of physics?). Sixty Symbols videos are instead a conversation with physicists about particular topics that interest and enthuse them – they represent a taster, rather than a tutorial. (I discussed my qualms about YouTube edutainment in a physicsfocus post last year.)

Leaving those points aside, TheRumpus’s comment raises a much broader and rather more subtle issue. Is adding to confusion necessarily a bad thing? Should we always avoid the possibility of confusing the audience when we’re discussing or explaining physics? Or could confusion actually aid the learning process?

That certainly seems like a rather, errmm, unhinged set of statements for a university lecturer to make. After all, don’t I aim to make my lectures as clear as possible so as to enhance student learning? Don’t I revise and re-revise the notes I give students in an attempt to eliminate any hint of ambiguity? And isn’t the quality of my teaching assessed (via, for example, Student Evaluation of Teaching questionnaires) on the basis of its clarity?

Yes to all three questions. But could this focus on eliminating confusion and ambiguity actually be doing students a disservice?

A fascinating article was published in The Chronicle of Higher Education back in August on exactly this topic. In “Confuse Students to Help Them Learn”, the work of Derek Muller, of Veritasium fame, and both Sidney D’Mello (University of Notre Dame) and Arthur Graesser (University of Memphis) on the role of confusion in learning is described. D’Mello and Graesser’s work challenges much of the received wisdom about teaching and learning and I’ve made time over the past couple of weeks to read a number of their publications (which are all available here).

The title of a paper published earlier this year by D’Mello, Graesser and colleagues nails their colours to the mast: “Confusion can be beneficial for learning”. The abstract does a very good job of bringing out the key points of their study. Here’s an extract:

“Confusion is expected to be more the norm than the exception during complex learning tasks. Moreover, on these tasks, confusion is likely to promote learning at deeper levels of comprehension under appropriate conditions”

This flies in the face of everything we’re told about the characteristics of effective teaching, but, I suspect, will nonetheless chime with many physicists’ experience of how they came to understand complicated concepts in, for example, quantum theory, relativity, and – oh, let’s say – thermodynamics and statistical mechanics.

Over a decade before D’Mello et al.’s paper was published, Kurt VanLehn and co-workers had found that in order for successful learning to take place in physics, an ‘impasse’ (as they describe it) has to be reached. In other words, the student must be confused at some point in order to learn.

Or, as Derek Muller puts it in that Chronicle of Higher Education article:

It seems that, if you just present the correct information, five things happen. One, students think they know it. Two, they don’t pay their utmost attention. Three, they don’t recognize that what was presented differs from what they were already thinking. Four, they don’t learn a thing. And five, perhaps most troublingly, they get more confident in the ideas they were thinking before.”

So, “add[ing] to the confusion about entropy”, as TheRumpus puts it, need not necessarily be a bad thing. What is of key importance, of course, is the student reaction to that confusion. We need to be very careful to ensure that the learner does not switch off entirely (and D’Mello and Graesser are at pains to stress this).

But when confusion triggers a response like the following, it’s difficult to argue that we should always aim for maximum clarity:


“Time to do some reading on this.” What more does any teacher want to hear?