In the first part of this article series, I discussed general guidelines for analyzing the results of scientific experiments (Klaus, 2015). The next step is the graphical representation of the results. The importance of data presentation should not be underestimated as figures are a key component of both data analysis and publications. Graphics are an important analytical tool as they can help to reveal patterns and illustrate differences. The appropriate plotting of data can also strengthen or even replace more formal statistical procedures, such as hypothesis tests. In the context of scientific publications, figures should guide the reader through the article and provide a clear and precise representation of the experimental results.
In this article, I will focus primarily on key principles and good practices for presenting small‐to‐medium datasets with the aim of comparing results from different experimental groups. As a general rule, authors should show as much of the actual data as possible instead of summarizing datasets via means or variances. Even larger datasets can be displayed efficiently using an appropriate plot; bars and boxes to visualize summary statistics can serve as additional visual guides. To adapt the methods described in this article, readers can download a supplementary “notebook” (see Code EV1) with code to generate the plots in the R language (R Core Team, 2015). Additionally, this web tool (http://embojserver.embl.de) generates the bee swarm plots and dot plots discussed later in the article. Apart from the topics discussed in this article, there are many more aspects that require attention. The “Scientific Figure Design Course” material by the Bioinformatics unit of the Babraham Institute (Babraham Bioinformatics, 2015) and the book by Tufte (1983) are valuable references.
I start with a discussion of displaying small‐scale experimental datasets. Let us assume that we have a fluorescent marker for detecting a recombination event in …
Subscribers, please sign in with your username and password.