Adaptive educational games –adjusting the difficulty level to children’s performance in the game– can be promising to foster cognitive, non-cognitive, and efficiency outcomes. 

However, empirical evidence regarding the effectiveness of adaptive games is limited. First graders will be randomly enrolled in three conditions: an adaptive game condition, a non-adaptive game condition, and a non-gamified
learning environment. In the adaptive and non-adaptive condition, children will play during a period of six weeks resp. an adaptive and non-adaptive version of two games for early numerical skills (number sense game) and early reading skills (reading game).

In a first set of four experiments, physiological data will be validated to index non-cognitive learning outcomes. Next, a large scale study will examine the effectiveness of the adaptive and non-adaptive games. Cognitive learning outcomes will be investigated immediately after gameplay and delayed, and both in terms of the trained skills (near transfer) and of math and reading in general (far transfer). Noncognitive outcomes will include self-reported math and reading motivation and anxiety as well as physiological data tracked during gameplay. Efficiency outcomes will be assessed by tracking the training time needed to achieve the targeted skills. Finally, we will investigate the role of student-level and environmental factors in the effectiveness of adaptive and non-adaptive educational games.