Online Program

Return to main conference page

All Times EDT

Friday, June 5
Data Visualization
User Testing Statistical Graphics
Fri, Jun 5, 3:30 PM - 5:05 PM
TBD
 

Scenarios of Visual Inference (308164)

Presentation

*Heike Hofmann, Iowa State University 
Susan Vanderplas, Iowa State University 

Lineups allow us to assess the validity and significance of a graphical result, and quantify its strength by assigning it a p-value with the same rigor of traditional statistical inference. We will discuss some examples of visual inference, present a theoretical model for the visual inference process, and demonstrate its flexibility in different experimental scenarios. Using a common framework, we discuss the theoretical and practical effects of different assumptions which are made when constructing a lineup plot and demonstrate the effects of these scenarios and assumptions on the calculation of visual p-values in experimental settings.