Christian Ritter, ISBA

Abstract:

Detecting and understanding structure and relationships in data represents a learning process, that is a way of constructing usable concepts in the long term memory of the concerned individuals. Statistical graphs and tables are two key interfaces with which the observer learns about structures and relationships in data. Making sense of statistical graphs and tables therefore involves a process of human perception. This talk presents a minimal theory for this process and uses it to explain why some statistical graphs and tables work well and others don't and why this also depends on the back ground of the observer.

(note: this is an adapted version of a talk given annually, so if you have seen it already you may not learn much new stuff)