Following just one of these "10 Simple Rules for Better Figures" will make you a better scientist

Communicating data in the form of a figure belongs to the highlights of scientific work. Posters, papers and presentations can greatly benefit from carefully crafted figures.  Nicholas Rougier et al. have devised a highly useful collection of ‘Ten Simple Rules for Better Figures’ (they must have been tempted to call these the 10 Commandments). Here they are:

1. Know Your Audience
2. Identify Your Message
3. Adapt the Figure to the Support Medium
4. Captions Are Not Optional
5. Do Not Trust the Defaults
6. Use Color Effectively
7. Do Not Mislead the Reader
8. Avoid “Chartjunk”
9. Message Trumps Beauty
10. Get the Right Tool 

In a somewhat self-referential twist, the paper uses eight figures to illustrate the rules. While I do agree with the comments on most figures, I’m at odds with one of them. Tell me if you agree.

Fig 1 In ‘Know your audience’ Nicholas et al. make the point that this depiction is OK for publication in the New York times but “…not acceptable in scientific publications if actual numerical values were not given elsewhere in the article.”. I do agree with this statement, a supplemental table in the respective publication would help fellow researchers to re-use the data. Unfortunately this is not standard practice.

Fig3 In ‘Adapt the Figure to Support Medium’ the authors make an excellent point for differentiating between figures show in print vs. slide in a talk. Yes, the right image is much simpler to grasp, isn’t it?

Fig 4 In ‘Do not trust the defaults’, a valid point is made in favor of “tweaking the various available settings” rather than using the out of the box axis labels. I agree, quite an improvement on the right.

Fig 5 Not sure if it is indeed “easier to see details in the high frequency domain” on the sequential colormap on the right.

Fig 6 This one I like best since it illustrates how the human brain can be mislead. Area and radius expressed as circles can suppress actual differences, whereas bars show them clearly.

Fig7 In “Avoid chartjunk” the message is clear: please make it easy for the viewers to correlate data sets with their depiction in a graph

.

Fig 8 A funny one. Who would have thought that Xkcd sketched look would make it into a PLOS paper?

Follow these Ten Commandments and your message will be understood better.

References:

Rougier NP, Droettboom M, & Bourne PE (2014). Ten simple rules for better figures. PLoS computational biology, 10 (9) PMID: 25210732