Scatterplots are not used by the NY Times because readers simply can’t make sense of them. Don’t oversimplify, but don’t assume that your audience can read a complex chart. Know your audience, and if possible test your charts with a small sample. Know what they expect, deliver that and perhaps a little more. When adding new charts, explain them thoroughly but make sure you have a backup plan (some pies…).
You may also be interested in:
{ 4 comments… read them below or add one }
I clicked on the link expecting something that would back up your assertion. Instead I found simply someone who shares your opinion. How disappointing.
FYI, I included in my blog a practical example how you can utilize scatter plot (XY chart) on sales data.
http://jpbi.blogspot.com/2007/11/viewing-sales-data-by-using-xy-charts.html
I totally agree with you: when adding new charts, (be ready to) explain them thoroughly.
Derek, virtually all authors emphasize the role of the audience, except perhaps Tufte and Bertin. There is an interesting article by Charles Kostelnick, “Conflicting Standards for Designing Data Displays” that could interest you (I don’t find a free link).
Janne, thanks for the link. You know, when you look at product sales data, growth and market share are probably the most used measures in the corporate sector, and they should be displayed in a scatterplot. But because people can’t make sense of them they are replaced by column charts.