Do Individual Differences Modulate the Effect of Agency on Learning Outcomes with a Serious Game

In order to support and promote new ways of learning, educational technology should be based on sophisticated theories and models of learning. Many issues are raised in the current understanding of learning by the constant evolution of educational technology and the burgeoning of educational contexts using these technologies. By examining the relation between agency and learning gains using a Serious Game for learning Physics, the present study focuses on a main issue of technology use: whether actively playing the game or watching someone play is beneficial for learning. Thirty-seven dyads participated in the study. Randomly assigned, one participant played a Serious Educational Game for learning Physics, Mecanika (Boucher-Genesse et al. 2011), for 120 min, while the other participant watched the player’s gameplay in real-time on a separate screen. As pretest and posttest, the Force Concept Inventory (FCI; Hestenes et al. 1992) was administered to measure learning gains in Physics. Analyses of answers on the FCI demonstrate that a Serious Game, such as Mecanika, is beneficial to learning, regardless if learning is conceived as relatively coarse shifts from wrong to good answers (scientific conceptions) or as more nuanced shifts from fillers/misconceptions to scientific conceptions. Also, individual differences in learning gains across dyads were found, which can be explained by the gameplay of a dyad created by the active player. Furthermore, the effect of agency is systematic and not modulated by individual differences: watchers learn more than players. These results need to be further explained by modeling the learning process.