Saturday, September 8, 2007

Needs Salt (Tufte)

In dealing with data representation, the Tufte piece makes strong points about mapping methodology and technique. He argues that the data maps are best when they are concise. He supports the idea that the focus should be content, not methodology or technique. His criticisms primarily relate to the lack of normalized data in the maps. In his discussion of the John Snow cholera map, he suggests the importance of discovery, and in the galaxy map discusses the possibilities that have been opened by mapping.

The maps he uses to illustrate his points attempt to display data in an unbiased manner, and Tufte seems to praise the effort. The flaws he identifies deal with statistical or representational inaccuracies. However, as we have discussed (but perhaps not agreed), no map is completely factual or unbiased. I also believe that calling attention to methodology and technique is an interesting way to open up a dataset or map, or statistic for that matter, to alternative interpretations.

No comments: