Learning Analytics for Educational Game Design

Big data in education (c.f. U.S. Department of Education, 2012) has fostered emergent fields like educational data mining (Baker & Yacef, 2009; Romero & Ventura, 2010) and learning analytics (Siemens & Long, 2011). Simulations and educational videogames are obvious candidates for the application of these analytic methods (Gee, 2003; Steinkuehler, Barab, & Squire, 2012), affording big data situated in meaningful learning contexts (Mislevy, 2011; Shute, 2011; Clark et al., 2012). In the design of these game environments, experts assert that players rarely interact with the game in exactly the way the designers envision, and thus heavily emphasize early, repeated usertesting (Schell, 2008; Salen & Zimmerman, 2004). With the added element of content-specific learning goals, or concrete growth over time in a domain-specific skill, attending and adjusting to organic play patterns becomes even more vital (c.f. Shute, 2011; Norton, 2008; Institute of Play, 2013). Thus, just as design-based research in the learning sciences involves data-driven, iterative refinement of measurement tools and experimental design (Barab & Squire, 2004), educational game design needs to incorporate learning-specific assessment mechanisms and leverage sophisticated techniques to understand nuanced learner patterns in play – informing development from the earliest stages of design.