A Data-Driven, Player-Centric Approach to Evaluating Spatial Skill Training Games

Abstract

Certain commercial games train spatial skills, a subset of cognitive skills used in school, the workplace, and everyday life. However, it is difficult to design spatial skill training games without knowing why some games are effective. In addition, existing game training studies do not analyze motivational factors for critical target populations. We conducted a study comparing spatial skill training effects and enjoyment for low spatial skill students on three training interventions: the computer game Homeworld Bound: Redux, digital spatial workbook exercises, and an active control group. We found no training effects for any intervention, but performance in certain game levels predicted spatial skill, and low spatial skill students enjoyed the game more than the workbook exercises. We provide three design recommendations for spatial skill training games based on our findings: use asymmetry for object manipulation tasks, require explicit 2D to 3D representation translation, and employ time pressure for navigation tasks.