Brynn Marie Evans

I came across this data visualization today. It’s really just a plot, but it conveys a lot of information in a very simple, intuitive manner. And I’ve been especially sensitive to this lately since I’ve been working on some information-rich visualizations of my own data.

Here is the plot:

It’s from a paper on gender identity in girls with Congenital Adrenal Hyperplasia (CAH), which has been thought to lead to male-typical behaviors. The study looked at whether girls tended to “feel” more male or female (scale from 0-18): normal girls, girls with CAH, and “tomboys” (as identified by their parents). To be honest, the outcome of the study is less interesting to me than this plot.

Each dot represents one girl’s gender identity score. Horizontal lines represent group means; vertical lines represent standard deviation. I really like how you can get a sense for how many subjects were in each group, how individuals compared relative to their group, and how the groups compared relative to each other. For example, it’s obvious that tomboys had the greatest variability in scores (and small N within the group), or that on average, girls with CAH scored in between control girls and tomboys—and that the average is likely not due to only a few outliers.

Overall, this is a nice, clean, information-rich presentation of data. Now I just need to figure out which plotting tool they used!