The impact of trace data and motivational self-reports in a game-based learning environment

Game-based learning environments (GBLEs) are increasingly present in school settings, primarily because they have been associated with improved academic and motivational outcomes. However, recent research found that GBLEs were no more motivational than conventional instructional methods. Yet, the disproportionate use of self-report measures may not be accurately representing children's motivation in these contexts. For this reason, this study sought to compare fifth-grade children's motivational self-reports with trace measures in their ability to predict both performance and content learning in Crystal Island Uncharted Discovery, a science-based GBLE. Self-report measures included achievement goals and situational interest; and, trace data included two measures of in-game engagement and one measure of disengagement behaviors. Results revealed that self-reported achievement goals and interest failed to significantly predict off-task behavior and accounted for only 3% and 5% of the variance, respectively, in predicting the two measures of in-game engagement. Most importantly, self-reported achievement goals and interest were not as effective as in-game trace behaviors in predicting performance and science content learning. Findings reinforce the importance of identifying and utilizing trace measures that represent indicators of performance and learning, while also considering the limitations of using self-report measures with children in GBLEs.