Visualization of data and findings are increasingly as important as the analysis techniques themselves. This can be a good thing when you consider that the goal of research is to do something with it, and data visualization techniques can foster understanding that facilitates doing. I have gone to Tufte trainings and found them interesting, and they even prompted some ideas for some information presentations for clients. However, Tufte’s examples (while thought-provoking and often beautiful) are not terribly pragmatic for the type of interview, survey, and experimental data I collect for my clients. Consequently, I have been looking for the last year and a half for a good, basic reference book that deals with different research visualization techniques. This is a review of two books I have tried.
The first book is a book called Diagrams: Innovative solutions for graphic designers by Carolyn Knight and Jessica Glazer I had extremely high hopes for. I had read great reviews, the cover had an interesting wordle-type graph on the front, and it came with a CD that came with “Copyright-free icons: arrows, flags, trees people and more.” In many ways this is a beautiful book. The graphics inside remind me of an interesting art show you may see in a hip gallery somewhere. Unfortunately, I should have heeded the title a bit more – this is a book squarely designed for graphic designers. I got a few ideas for some research visualizations, and the CD had a few icons I may use, but research is definitely not the key focus and this is not what I was looking for in terms of a solid reference.
After my experience with Diagrams, I decided to get more specific. I looked for a book that dealt with visualizations of quantitative data, and I was turned onto Stephen Few’s, Now you see it: Simple visualization techniques for quantitative analysis. The title alone had me hopeful and a quick perusal of the book lifted my heart as I saw some familiar graphs and charts. Few’s book is actually quite good, but I would say his treatment and breadth of visualizations are (for lack of a better word) clinical. There are many scatter plots, bar chars, and line graphs. This book would be a nice compliment to a basic statistics book. It shows how scatter or line plots change as a function of different relationships within data. I also like that Few presented some commonly used software packages for visualizations and presented some good evaluations and examples of these packages.
All in all, Few’s book is the closest I have found to a reference guide on visualization, but many of the techniques and visualizations were not terribly innovative, and maybe that is OK. However, I am still hoping that someone will publish a book somewhere in between these two. A quantitative focus, grounding and ease of use in Now you see it, with the boundary pushing and new ideas in Diagrams. If anyone knows these authors, please introduce them to each other. Until then, this researcher is still dreaming of the perfect companion to help visualize data.