Does a ‘good story’ get in the way of the truth?

Humans are storytellers. Before there was recorded history there were oral traditions, some of which – such as Homer's Iliad and Odyssey – endured long enough to enter into written language and then print. And part of what enabled that oral tradition was story – the use of narrative forms that both brought to life what was being conveyed, and made it easier to remember and pass on.

So it's hardly surprising that academic researchers – who are excited about and interested in their topic – often convey what they are searching for and what they have found as a story. In fact, early career researchers are often told they need to make sure that their data 'tell a good story'. But is this focus on telling a story a good thing, a bad thing, or a bit of both? How much is it our job as psychologists and researchers to change minds?

Informing and persuading

Discussions around incentive structures in academia are well rehearsed – we collectively place too much weight on some things, not enough on others. Critically, we perhaps prioritise the wrong things. Part of this includes our enthusiasm for positive findings; a decade ago, Annie Franco and others wrote in Science on how researchers are less keen to write up their own null results, and I have written with others on the fact that our peers are less likely to cite them, which impacts the apparent strength of evidence for treatments.

Does our enthusiasm for telling a good story feed into this? Does it encourage the use of Questionable Research Practices (QRPs) – the search for ever smaller P-values that will help us tell a more convincing story, that will be more likely to be published, published in a certain kind of journal, and so on? And does it fuel hype in the media? There is certainly evidence, from Petroc Sumner – with Chris Chambers and others – that much of this comes from academics themselves, through academic press releases.

Part of this turns on the question of what we are trying to do through scholarly communication. Are we merely trying to inform our peers as to what we did and what we found? Or are we trying to persuade them that what we have found is interesting, and important, and should change how we view the world? The reality is that it is both. But are we always sufficiently clear about what we are doing and when?

The distinction between informing and persuading is well understood in some quarters – by journalists and documentary makers, for example. But whilst some professions (such as journalism) offer formal training in the distinction between communication intended to inform vs persuade, academia typically does not. The UK Reproducibility Network (UKRN) recently ran a workshop on informing vs persuading in scholarly communication to highlight this distinction, why it matters, and to discuss the role of researchers, journal editors and others in helping draw out these different aspects.

In practice, of course, there will be elements of both informing and persuading in any actual piece of communication. But what elements of persuasion can be controversial? Maybe the selective use of data, selective summaries of the balance of existing evidence, moving from a finding to a 'call to action', the elision from association to causation, use of narrative as a persuasive device, and/or minimising uncertainty.

Many of these elements are already considered QRPs. But what underlies them is often a desire to persuade – either of the importance of a finding, or of its truth and implications. Can that desire be corrupting? And how do current incentive structures in academia – e.g. pressure to describe our research as original, rigorous and significant or novel) – contribute to the elision between informing and persuading?

Should we be thinking more about this distinction, and the difference between the two? What constitutes persuasive communication in an academic context, and how does it compare to informative communication? Can a particular piece of communication seek to both inform and persuade, without undermining research integrity? All of this was discussed in the UKRN workshop.

Full, honest and transparent

People will have different views on whether or not storytelling has a place in scholarly communication. Some will say no, others will say it is at the heart of good communication, and that should include scholarly communication. But storytelling and persuading are not necessarily the same thing. Can we focus on (primarily) informing and tell a good story?

If we accept that storytelling is an inevitable part of all communication, then we need to make sure that the desire to tell a good story does not get in the way of giving a full and honest account of our work. Open research practices – making as much of our research process and intermediate research objects as available as possible – can perhaps help with this.

By all means, tell a good story. Attempt to persuade us that we should take note of what you have discovered (or think you have discovered!) if you think that's appropriate. But also let us check your working.

How to tell a good scientific story

So, if you do decide you want to tell a good story, how do you go about it?

Social psychologist Mick Billig has written about how the person has been removed from the psychological literature. Even though scientific enquiry is often a very personal endeavour – we study the things we are interested in – we are taught to communicate what we do and what we find in a way that often acts against this. We are advised, more often than not, to write in impersonal (and ostensibly objective and disinterested) way – third person, passive voice, and so on. How often do you read something along the lines of 'This study found…' – when we were the ones doing the finding!

We don't have to write this way. A slightly more personal style – first person, active voice – can be much more engaging to read and more honest, making it clear that it was us, as motivated and fallible humans, that were doing the doing. In other words, perhaps it is easier to be transparent and honest about what we did, how we did it, and why, if we write in a more natural way.

Some classic studies in psychology do exactly this. Take The Magical Number Seven, Plus or Minus Two by George Miller. Written in 1956, this breaks pretty much all of the rules we teach students about scientific writing! But it's a lovely read. Simultaneously very personal and very honest – presenting data, drawing conclusions from those data, and being careful not to go beyond the data (as much as possible).

So perhaps the 'rules' of scientific writing we are taught are actually part of the problem. They impose a structure intended to convey a sense of objectivity and impersonality that does not accurately reflect how science actually works. The curiosity, the exploration, and the uncertainty. Writing in a more personal way could be both more engaging and more honest, and support wider efforts to be transparent.

Marcus Munafò is Professor of Biological Psychology and MRC Investigator, and Associate Pro Vice-Chancellor - Research Culture, at the University of Bristol. 
[email protected]

Editor Dr Jon Sutton comments:

You may have noticed, over the years, that I bang on about storytelling in scientific communication a lot. This edition is filled with pieces that discuss creative ways of telling a story around research in Psychology. As a magazine, I constantly stress to contributors that we are looking for 'Personal, Practical, Persuasive'. But it's starting to feel like I'm ploughing a lonely furrow.

When I give talks, I use the Miller example which Marcus includes below, and the intro from the Dunning-Kruger paper ('But I wore the lemon juice!' - see illustration above, by Gemma Hill). But say you, as a journal reviewer or editor, got a submission in Miller's style. Would you simply go 'What on earth is this?' Is that ever going to change?

Even with publications that lean more towards 'magazine', I'm not sure everyone is as keen on 'story' as I am. Recently I was talking with Aisling Irwin, freelance science journalist and Association of British Science Writers board member, and she said: 'There's an aspiration on the part of feature-writing science journalists, fed by what they read in mainstream publications, to write character-driven narratives – but it's simply not wanted by their editors. I find, talking to the members of our science writing group, that they often have sought out interesting colour and personal stories that don't make it into the final version.

So, my thinking had been that editors of specialist publications are after something different – something more pragmatic, not meandering around but delivering facts for busy people (while of course being easy to read).'

Of course, I still feel that the best writing is pragmatic and practical, and driven rather than meandering. But ultimately it's starting to seem to me that this all comes back to the presence of the 'personal' and 'persuasive' in academic writing. I get that people are often wary of this, conflating persuasion with confabulation. I honestly don't think it has to be that way: but it's only the gatekeepers of tomorrow, usually the early career researchers of today, who are going to truly drive change.