The purpose of this study is to explore some of the ways in which gameplay data can be analyzed to yield results that feed back into the learning ecosystem. There is a solid research base showing the positive impact that games can have on learning, and useful methods in educational data mining. However, there is still much to be explored in terms of what the results of gameplay data analysis can tell stakeholders and how those results can be used to improve learning. As one step toward addressing this, researchers in this study collected back-end data from high school students as they played an MMOG called The Radix Endeavor. Data from a specific genetics quest in the game were analyzed by using data mining techniques including the classification tree method. These techniques were used to examine the relationship between tool use and quest completion, how use of certain tools may influence content-related game choices, and the multiple pathways available to players in the game. The study identified that in this quest use of the trait examiner tool was most likely to lead to success, though a greater number of trait decoder tool uses could also lead to success, perhaps because in those cases players solving problems about genetic traits at an earlier point. These results also demonstrate the multiple strategies available to Radix players that provide different pathways to quest completion. Given these methods of analysis and quest-specific results, the study applies the findings to suggest ways to validate and refine the game design, and to provide useful feedback to students and teachers. The study suggests ways that analysis of gameplay data can be part of a feedback loop to improve a digital learning experience.
All Science Journal Classification (ASJC) codes
- Computer Science(all)