Design and Development of Visualization Approaches for Informal Learning Game Logs

This study addresses the "observation problem" in informal game-based learning, specifically the challenge of understanding how players' behaviors relate to their learning progress within open-ended educational games. Focusing on Hist Maker, a mobile puzzle game about history, the authors argue that prior cluster analysis of 185 players, which identified three behavior groups (Explore, Hint, and Negative), was limited to a small number of parameters and could not reveal how behavior shifts as exploration advances. The work proposes and evaluates visualization approaches designed to link player actions to exploration progress, measured by the number of historical formulas acquired.

The first approach applies line charts and heat maps to compare average action frequencies across the cluster-derived player groups. Its findings indicate that trial-and-error behavior and reliance on database tools increase in later phases of progress, and that two averaging methods yield complementary views of group behavior. The second approach adapts the Wisteria Graph, a node-link representation, to display individual players' time-sequenced actions, allowing researchers to infer gameplay context that aggregate analysis obscures.

Evaluation through interviews with five educational technology researchers found unanimous agreement that both tools clarified the relationship between behavior and progress. Participants recommended combining action-frequency charts, reducing visual clutter, and adding contextual information. The authors conclude that group-level visualization provides representativeness, whereas individual-level visualization offers richer detail at the cost of generalizability and readability. They therefore recommend integrating both approaches to support more complete assessment of informal learning in games.