Seeing is Believing

Abstract

Big data sets can be cumbersome and difficult to understand. User-centered and interactive graphical displays help communicate messages from large and complex data as well as provide a new method to identify data trends outside of tabular or statistical analysis. Human factors researchers can utilize data visuals to not only develop but also answer questions that previously proved difficult through visual exploration. This approach is especially relevant to the field of surface transportation research where complex plots can incorporate both temporal and geospatial data in an easy-to-digest format. As a proof of concept, this paper demonstrates how bike-sharing and historical bicycle collision incidents can meaningfully merge to produce graphical displays that readily identify and communicate potential infrastructure problems for safety. Through the use of Bayesian modeling and geospatial mapping, our analysis identifies two primary trends worth further consideration and research to consider for cyclist safety in Chicago.