John Wilkins at Evolving Thoughts has advice on how not to get your Cell paper retracted for improperly massaging images. This just happened to a group from Taiwan, and I don’t know the whole story, but Science is supposed to have an article about it Friday.
Now that most data are entirely digital from collection through publication, tweaking has become way too easy. I love Photoshop, but it’s important to recognize the difference between making a figure effective, and making the data themselves stronger. John’s rules are good ones.But I’d disagree with his statement that
Aesthetics do not matter in scientific papers; honesty does.
Yes, honesty matters more. But the most effective papers have pleasing figures. Aesthetics help make papers effective. There are a myriad of ways to present data, and the habitual mode of presentation within a research group is often not so clear to outsiders. Simple aesthetic choices — such as logical orientation of specimens, avoiding extreme magnifications in which the larger context of the data is lost, aligning control data with experimental data in a regular, intuitive manner — can have a huge impact on clarity. Reading a paper with badly composed, cramped figures is like slogging through wet sand.
I consider aesthetics every time I walk my undergraduates through a journal article. Students are spoiled by airbrushed textbook diagrams, and when they first look at actual data, they can be totally baffled. I’ve had students who mistook blots for cells, cilia for hair, heads for tails, etc. The worse the figures are, the more discouraged the students become. But if the figures are attractive, almost everyone will read the entire paper.When I choose articles for my syllabi, elegance of presentation counts as much as scientific significance. It’s simple: the clearer the figures are, the more the students understand and retain. In fact, I wish figure design was part of the graduate curriculum. But that’s another lament entirely. . .