What happens if you like a study, but find the way that the authors have chosen to present their data unintuitive or unhelpful (or even just not pretty enough)? Last month we published an elegant Research Article by Ram Maharjan and Thomas Ferenci about the surprisingly diverse effects of different environmental stresses on the spectrum of mutations arising in bacterial populations (you might also want to read this excellent Primer by Deepa Agashe). When tweeting the paper from @PLOSBiology I decided to illustrate it using the authors’ Fig 5, which I thought summarised the papers’ main claim reasonably well. Clément Viguier, a PhD student at the University of Grenoble who’s interested in data visualisation, agreed that the study was intriguing but wasn’t so sure about the merits of the image I’d picked:
Now, one could indeed debate the merits of the authors’ 3D graph as a way of encapsulating a moderately complex set of data, but luckily, because of PLOS’ rigorous data policy, there was an easy reply here; we insist that the authors should make the numerical data underlying each of the main and supplementary figures available to readers, either as supplementary data files or as depositions in an online repository. I pointed this out to Viguier, and thought that would be the end of the story, but a few weeks later, he replied:
You can read Viguier’s own side of the story in his blog post here, where he discusses the challenges of putting his money where his mouth was, and the pros and cons of different approaches to displaying these data; you can also see the range of visualisations that he tried and obtain the code that he used. Tom Ferenci – senior author of the original paper – said “I admire the passion and also admire the bubble chart that he came up with – it is visually the prettiest. On the other hand, the problem with all the heat map approaches is that they do not display the relative magnitudes in an easy-to-appreciate way.” They had actually experimented with heatmaps themselves but felt the 3D representation worked better.
“One element missing from the discussion is that there is a strong element of ‘taste’ in what visual representation best works for a viewer,” continued Ferenci. “Why choose an Audi over a BMW except styling and personal taste? Clément obviously hates 3D representations but the defects he presents are minor relative to the boring lack of impression the flat heat maps generate – that is why we went with the more dramatic graphical landscape version.”
There are clearly many ways to skin a cat, as the unfortunate English adage has it, and perhaps one day all journals will present the data in an immediately implementable form; readers who like heatmaps could click “heatmap” and those who prefer 3D could click “3D,” etc., but at least the provision of the underlying data meant that this feline flaying was even a possibility. Long live open data!
Featured image credit: Clément Viguier