Impact of time-based visualization on situation awareness

We have developed a time-based visualization system and tested it in accordance with a well-developed theory of situation awareness during a simulation of an actual emergency. The client-server system is fed by a collaborative server, and uses linguistic parsing, suffix tree clustering, and multidimensional scaling to develop a changing picture of textual data related to an unfolding event. We provide a detailed account of situation awareness theory and use it to explain the effect of the visualization on perceptions of facts, comprehension of the situation, and projection of near term status. We developed a simulation of a tragic shooting at the University of Arizona, performing requirements elicitation of officials involved in the actual event, and preparing a repeatable scenario based on, but differing from, the event. We ran the simulation with about 200 management students, half of whom used the visualization system and half of whom used a competing text-based system, and found that the visualization system aided comprehension and projection, but had no effect on perception of facts.