Games can be rich environments for learning and can elicit evidence of students’ conceptual understanding and inquiry processes. Illuminating students’ content-specific gameplay decisions, or methods of completing game tasks related to a certain domain, requires a context that is open-ended enough for students to make choices that demonstrate their thinking. Doing this also requires rich log data and methods of Game Learning Analytics (GLA) that are granular enough to look at the specific choices most relevant to that context and domain. This paper presents research done on student exploration of high school level Mendelian genetics in a multiplayer online game called The Radix Endeavor. The study uses three approaches to identify content-specific gameplay decisions and distinguish players utilizing different methods, looking at actions and tool use, play patterns and player types, and tool input patterns. In the context of the selected game quest, the three approaches were found to yield insights into different ways that students complete tasks in genetics, suggesting the potential for a set of more generalized guiding questions in the GLA field that could be adopted by learning games designers and data scientists to convey information about content-specific gameplay decisions in learning games.

游戏可以是一个丰富的学习环境，并且可以为学生的概念理解和探究过程提供证据。要阐明学生针对特定内容的游戏玩法决策或完成与特定领域相关的游戏任务的方法，就需要一个开放的环境，以使学生能够做出能够证明自己思想的选择。要做到这一点，还需要丰富的日志数据和Game Learning Analytics（GLA）的方法，这些数据和粒度必须足够精细，以查看与该上下文和域最相关的特定选择。本文介绍了一项在名为Radix Endeavor的多人在线游戏中对高中水平孟德尔遗传学的学生探索的研究成果。。该研究使用三种方法来识别特定于内容的游戏玩法决策，并使用不同的方法来区分玩家，着眼于动作和工具使用，玩法和玩家类型以及工具输入模式。在选定的游戏任务中，发现了这三种方法可以洞悉学生完成遗传学任务的不同方式，这表明在GLA领域可能出现一系列更广泛的指导性问题，学习型游戏设计师可以采用这些指导性问题和数据科学家在学习游戏中传达有关特定内容的游戏决策的信息。