Tuesday, September 6, 2011

Reaction: Graphical Perception - Theory, Experimentation, and Application to the Development of Graphical Methods

This article is an early take on defining what are the most effective graphical representations of data. It breaks down how people extract quantitative information from graphs into basic elements and then ranks them by effectiveness. The authors talk about an experiment they did to test their hypothesis on the ordering for the elementary perceptual tasks. I was wondering how effective it would really be to approach the problem in this way. Instead of having an open experiment to see what tasks were easiest, they seemed to be just trying to prove a very specific hypothesis.

I think this article did a good job of breaking down some of the differences between tasks people perform when looking at data. I'm not sure the tests that they were conducting were truly useful, though. I always think of graphs as being used to express very broad ideas or general principles, but they were measuring very precise data. From an earlier paper that I read, those authors seemed to say that when you wanted to convey precision, you shoud display the data in a table. That seemed appropriate here.