快播成人

Faculty, staff or current student? Take the UW Climate Survey

Researchers' 'diamond plot' data visualization technique featured in Nature, Fast Company

Scatterplot graphs
A scatterplot graph with a traditional X and Y axis (top) and a "diamond plot" (bottom).

Could simply tilting a scatterplot graph鈥檚 axis help prevent people from confusing correlation and causation?

Research by Information School Assistant Professor Jevin West and colleague , a biology professor and affiliate iSchool faculty member, proposes a 鈥渄iamond plot鈥 approach to data visualization, tilting the graphics at a 45-degree angle. At this angle, neither the X axis nor the Y axis appears to take priority or imply that one factor causes the other.

The research, first posted in a , was featured Oct. 4 in and in .

Simply tilting graphics won鈥檛 solve the fact that the media, the public and even scientists themselves often read causation into data when they are simply seeing correlation, but West and Bergstrom theorize that it could help address the persistent problem.

鈥淵ou could go through the last 2 weeks and you could find any number of examples of cases,鈥 Bergstrom told Nature. 鈥淩ed wine causes this, chocolate causes that.鈥

The duo, known for their course and curriculum, don鈥檛 make any such assumptions in their paper. Their next step is to investigate how people read the diamond plots through user testing.