In this paper, we propose a serious game for emergency preparedness that simulates a mass emergency caused by a terror attack in a train station and provides recommendations about
emergency evacuations. We evaluate the interactive vs. the non-interactive simulation of such scenario. In the interactive simulation, users actively play the game with controllers; in the
non-interactive simulation, they watch the same storyline progress automatically, showing them all the meaningful events which players must deal with for completing the corresponding interactive simulation. In other words, the non-interactive version can be considered as an instructional video made by recording video game action. In addition to assessing knowledge gain as other evaluations of serious games do, we measure users’ perceived self-efficacy (which has been shown to be a predictor of real-world performance) (Bandura, 1997) as well as other variables that are significantly linked with positive changes in protective behaviors (Floyd, Prentice-Dunn, & Rogers, 2000).