An Analysis of Game-Based Learning for the Design of Digital Science Games

Thesis (Ph.D.)--University of Washington, 2016-03As science becomes increasingly data-driven, there is a need to prepare the next generation of youth with the wide variety of skills and tools necessary for future scientific careers. In this dissertation, I address the diverse gameplay capabilities present in youth, arguing that educators and designers can and should leverage these in the design of educational science games. I employ both quantitative and qualitative approaches to examine gameplay within the context of the bioinformatics game, MAX5. Two initial studies are presented showing that a player’s previous experience with a game’s genre and the types of messages shared between players during social gameplay are significant predictors for learning outcomes. A qualitative data analysis then identifies themes of gameplay that are compared and contrasted with existing theories to lay the building blocks for a new complex systems model of game-based learning. Components of this model exist within five interlinked layers: the input, the sensory sphere, the structural dynamics, semiotic translation, and memory-action patterns, all existing within a larger dynamic network of games and players. These research findings provide a means for game designers to broaden the participation of youth in the sciences by matching player capabilities with appropriate game elements and learning content. This research further highlights the need for more adaptive science games that reflect not only players’ varied capabilities, but also the increasingly multidisciplinary and collaborative nature of scientific practice