Lord of the Files, part 2

Back in November, I wrote about the Diederik Stapel affair that was rocking the psychological community. Now, it’s all happening again – recently, Uri Simonsohn has found evidence suggesting that another Dutch social psychologist, Dirk Smeesters, has been tinkering with data to produce more desirable outcomes in his research.

We’re still waiting on Simonsohn’s paper on the matter to be published, so I think it’s a little too soon to speculate on how he came across Smeesters’ work – some have already started to make melodramatic comments likening Simonsohn’s approach to medieval torture. Clearly, we need to be very cautious about using such approaches as a sledgehammer. We don’t want to end up in a situation whereby completely innocent researchers wrongly getting caught out. But there are also some tough questions that need to be asked about how and why this sort of behaviour is happening.

One thing that hasn’t been mentioned all that much in the media hype around both the Stapel and Smeesters cases is the collateral damage caused by their misconduct. Here’s what I wrote on the matter about the Stapel case in November:

…I feel acutely sorry for the PhD students that Stapel, in supervising them in such an irresponsible way, has dragged through the mud. These students never got the chance to experience the full training that a PhD program is supposed to convey – they never got to run the experiments for themselves, to see how things go wrong, how to deal with setbacks. The few students or colleagues that did try to ask for access to the data were given excuses, and in some cases threatened or insulted. Gaining a PhD is a rite of passage in the academic world, and it’s nothing to do with getting some letters after your name or coming up with the next new game-changing theory. It’s about developing an inquiring mind, about gaining expertise incoming up with sensible, testable hypotheses and experiments, and learning how to accept that it’s okay when things don’t work. Sometimes a null result can be just as interesting as confirmation of a hypothesis.

I think this still stands, and now seems to be compounded by this latest scandal. It looks like a few people who worked with Stapel also worked with Smeesters, and are now having to go through another round of papers being retracted and reputations being smeared. Some, like Jonathan Levav, have started to speak up in defence of themselves and their colleagues, and it’s clear that that emotions are (understandably) running high (it might also be worth watching Camille Johnson’s blog, another of Smeesters’ colleagues, in the coming weeks to hear her perspective on things). For any PhD students involved, this is going to be particularly painful. For one, it means they’ve been let down in their training – one of the most important things that I learned during my PhD was that world is a big noisy, messy place, and that’s okay. Data aren’t always neat and tidy; in fact neat data are the exception, not the rule. I worry that a number of students that have gone through these labs will spend a good chunk of their careers with the expectation that results need to be pristine in order to be meaningful, and that hurts both them and the science involved.

Furthermore, the retractions could be potentially disasterous for their short-term career prospects. The pressure to publish is huge right now, and something that I myself am acutely aware of. Alongside the feeling of elation when a paper is published, there’s increasingly also the feeling of relief that it’s another one that you can add to your CV to make yourself look more employable. For Stapel and Smeesters’ students, having their work retracted will be an awful, awful blow.

So yes, Psychology (and perhaps science in general, as it’s happening in other areas) needs to take a long hard look at how it’s got into this mess, and make amendments sooner rather than later. Funding bodies like the EPSRC are going to be plugging huge amounts of money into Doctoral Training Centres in the coming years, and with this influx of PhD students comes the duty to train them in appropriate, ethical, and sustainable research practices – otherwise, the same mistakes will just be made all over again.