Learning to solve problems with animations

Observing animated models can be an effective instructional method for learning to solve abstract problems. In an animated model a pedagogical agent provides explanatory text for an animation in which a problem is solved. In three experiments with pre-university students (in Dutch: vwo 4) we investigated how such animated models for the domain of probability calculation could be further optimized. The results indicate that written explanations in animated models can be effective. It is important that students are enabled to take full advantage of the characteristics of written explanations. This can be realized by allowing students the control over continuous animated models and by stimulating reflection on animated models. In addition, the results reveal that the effectiveness of design guidelines is contingent on the alignment between the control that students expect and the control that they can actually employ. Theoretical as well as practical implications are discussed.