I am not a visually oriented person or, more appropriately, I am less than stellar at design. This may not be a surprise to anyone that has seen my attempts to assemble a wardrobe, but this is also true in the sense of organizing information - whether it is in a paper, on a poster, or for a conference presentation. As such, I am compelling myself to learn two programs/languages this summer that will both allow me to overcome my shortcomings ~ LaTeX and R. Yes, my devotion to wysiwyg interfaces and minimal .do files might be crumbling a bit. I haven't decided to convert quite yet, but am willing to pick up a second proficiency in the name of communicating my ideas. The study of LaTeX (which does not seem so daunting now) is purely driven to reduce my own fiddling with bad template design while the studying R is not intuitively related to design.
The Statistical Modeling, Causal Inference, and Social Science weblog linked to a host of graphs and figures that are designed for humorous intent. There are a few sites with similar graphs that can be googled, but I figured it was ample motivation to link the R - graphing gallery; those who already have a familiarity with R or this gallery are free to ignore tes link as I am sure several of you may have already come across it, but I figured there was enough non-R readers who may appreciate some of the graphs that are created with the program.
This Graph, for example, is a great way to show an evolving distribution over time with the normally relevant characters (central tendency, confidence intervals) and providing more information that does not complicate the chart too much, in my estimation. Some of the graphs are just visually appealing, though probably not as useful for a paper. While not an argument to switch over to R in itself, the examples provide some thoughts on how to reorganize my visual presentation of data. Additionally, I have started keeping track of both Flowing Data and Junk Charts weblogs. The latter seems to have a decent set of examples of what not to do or how to do things better. Given the relevant concern of our research being accessible to non-academics in some fashion, properly created graphs can be a mechanism to achieve this end.
A well constructed graph is worth its weight in abstracts...and then some.
Comments