When claims about science become edutainment, science loses

I think I understand the motivations. There are books to sell. Punters to please. Grants to get. Reputations to enhance. $45K keynotes to deliver. Teeth to re-veneer. That’s Edutainment! So nuance and humility – which should be hallmarks of science – aren’t going to cut it. Any disclaimer will make the declaimer’s field – and, by extension, themselves – seem weak and uncertain.

I also wonder sometimes if another reason for over-claiming is that it helps some scientists manage their own discomfort. Over the course of their careers they may realise, to their dismay, that Johnny Nash was probably right: there are more questions than answers, and the more we discover, the less we know.

What’s so bad about making the Claim? For a start, we don’t know anything from science because we can’t. We have only very partial knowledge. All we really can say is that given our limited data and the constraints of the methodology, a particular finding is more or less likely. And we can sometimes have a reasonable estimate of probabilities and likelihoods.

What about the claim that studies show something? Sure, some studies show something. But other studies do not show the same thing. So, again, it’s about probabilities rather than studies showing something.

On top of this, published studies tell only part of the story. In many fields, hypothesis-supporting positive results are much more likely than negative results to get published. So the claim that studies show refers only to the unrepresentative bunch with positive results that actually see the light of day. This publication bias represents another quite bizarre rejection of a scientific hallmark by its own practitioners: to publish all your results, not only those you like.

Also, the Claim fails to consider future research. Of course, we can’t know what this will find, but we do know it’s quite possible that it will reveal apparently well-established findings to be quite untrustworthy. New teams of researchers may fail to replicate the Claim. New research may even reveal that the methodology behind it – and behind perhaps thousands of other studies, carried out over decades – is flawed.

Isn’t it just wrong for scientists to behave in ways that violate the basic principles of science as a profession and endeavour? When pharmaceutical companies selectively publish only the positive results of drug trials, we are outraged. When car manufacturers find ways to distort the levels of emissions produced by their vehicles, we see it as corruption. Yet, somehow, when scientists do something similar by making overblown claims, we don’t make similar judgements.

But others may. Making overblown claims undermines trust in science and scientists because such claims, however confidently asserted, are very fragile and can easily be challenged or refuted. A single contradictory finding or questioning voice can be enough to shatter confidence, making it easy to interpret the Claim as a lie, the person who made it as a liar and science in general as fake news.

The possibly self-interested and certainly bland “more research is needed” conclusion of many scientific papers is quite wrong. We do not need more research. We need better research. This means improving our practices around conducting, publishing and communicating science.

But while scientists’ incentives remain as they are, it is hard to imagine them seeing the fun in that.

Rob Briner is professor of organisational psychology at Queen Mary University of London, a visiting professor at Oslo New University and associate director of research at the Corporate Research Forum.