Visualizations for the Assessment of Learning in Computer Games

Mounting evidence suggests that current trends in global energy usage are leading to global warming, which will likely change our climate irreparably. Yet, as noted in IPCC reports, most people do not take this danger seriously enough to change their behaviors. Computer games, which are increasingly being used for educational purposes, have the potential to change people's understanding and attitudes toward critical issues such as energy use and global climate change. Yet it remains unclear how well serious games achieve these ends, and what, exactly, it is that makes them effective. We propose that by looking at data collected by these games, and correlating it with instruments that measure changes in attitudes, we can determine what game scenarios and activities are actually changing people's minds. This will help us to design more effective games for educating the public in a way that yields tangible results. In this paper we describe a novel strategy for classifying and visualizing the dynamic, multivariate data generated by serious games. Our contribution is a framework for categorizing these data, corresponding to layered visualizations that help to reveal the patterns in what players are doing over time. Specifically, this paper introduces the concept of Action Shapes, which are glyphs that are automatically generated using a variation on parallel coordinates. As elements in the layered visualization, Action Shapes represent the "benificence" of students' choices seen in the contexts of student progress and the overall game state. As proof of concept, we are applying this visualization to Energy Choices, a multiplayer game that teaches people about the interrelated issues of global warming and energy use. Although the examples provided in this paper are specific to this particular game, this strategy may be readily applied to a wide variety of other educational games designed to help people to be smarter about energy use and the planet. Information visualization; learning assessment; serious games; glyph-based techniques; parallel coordinates