Investigating Insight Generation and Decision Making with Visualizations in Real and Virtual Environments

Visualizing data has been touted as a method to reduce cognitive workload by externalizing cognitive processes and utilizing the human perceptual system's ability to recognize patterns [1-4]. The current study investigated whether displaying sociocultural data in an immersive 3D virtual environment improved insight generation and decision making over traditional 2D visualizations. Participants were given 10 minutes to interact with visualizations representing sociocultural information and decide where a possible future threat in a fictional city may occur using either a 2D tabletop map in the real world or the same map represented in 3D in an immersive virtual environment. Visualizations included information either highly correlated with previous incidents (i.e., government building locations and bus routes) and information not highly correlated (i.e., religious centers, income for areas, park locations, and flood zones). Success required identifying which variables were highly correlated with previous incidents, and thus predicted where a future threat may occur. Initial results indicate that in the 3D virtual environment, successful participants gave their final decision about 90 seconds faster than unsuccessful participants. Throughout the simulation, previous incident and government building location visualizations were highly utilized (for both, 85% of total duration), however, successful participants used the bus routes longer (12%) compared to unsuccessful participants. Further, the current trend suggests that successful participants avoided using irrelevant visualization information compared to unsuccessful participants. Ongoing work will compare these results to the 2D table-top display.