Abstract:

Considering the recent influx of Game-Based Learning Environments (GBLEs) that seek to promote learning and achievement, there is a need to understand how certain in-game features can impact domain-specific achievement. This study compared the relative predictive power of off-task in-game behaviors and traditional self-reports of task-avoidant strategies across genders.

Grade 5 students’ (N = 797) played CRYSTAL ISLAND – UNCHARTED DISCOVERY, a science-based GBLE focused on landforms, map navigation, and map modeling. The game is a 3D immersive environment in which the player attempts to complete a series of six content-rich quests and one final culminating integrative quest. Students played the game in school over the course of five weeks. In-game off-task behaviors consisted of seductive details, such as sand dollars accumulated, sand dollars traded, and treasure chest time, which did not aid in solving game quests. A measure of science achievement was given after students played CRYSTAL ISLAND – UNCHARTED DISCOVERY for five weeks. Task-avoidant strategies, measured using the Achievement Goal Questionnaire, was operationalized using a combined score of performance-avoidance orientation and the mastery-avoidance orientation.

Results from multiple regression analysis indicated that sand dollars accumulated was the strongest predictor of science achievement, adversely impacting science achievement. In addition, off-task behaviors within the GBLE were stronger predictors of science achievement for boys than self-reported task-avoidant strategies. For girls, self-reported task-avoidant strategies and some in-game off-task behaviors were predictors of science achievement. These findings imply that both self-report measures of task-avoidance and in-game-measures of off-task behaviors predict science achievement.

Keywords:

Science Achievement, off-task behaviors, game-based learning environment.