Extracting Learning Analytics from Lightbot GameplaySessions

This paper presents a method for extracting and visualizing learning analytics from Lightbot using computer vision. Lightbot is a serious educational game that foments computational thinking and teaches introductory programming concepts. We collect information from recorded game sessions, such as the number of commands used, the effectiveness of the player’s solution, and the number of times the player submitted their solution for assessment. We then use this data to create graphs that help education professionals understand the challenges and successes students find when using Lightbot as a teaching tool.