Not the silver bullet: Learner-generated drawings make it difficult to understand broader spatiotemporal structures in complex animations

•An event unit analysis was applied to a complex animation.•The analysis served to identify increasingly broad spatiotemporal structures.•Drawing made it difficult to understand extensive spatiotemporal structures.•Drawing might let the learners focus on visuospatial aspects of an animation.•Demonstrating a physical model might be more effective than drawing.

Because drawing is a highly successful strategy in learning from text, it has recently been investigated whether drawing can also improve learning from animation. Several theoretical and practical arguments, however, make drawing a questionable strategy for learning from animation. In an experimental study, we investigated the effectiveness of drawing for learning from animation. One group of 26 students had to draw what they had observed in the animation. A second group of 26 students had to reflect on what they had observed in the animation. After learning, all students had to demonstrate their understanding by making use of a physical model. The students' demonstrations were assessed by means of an event unit analysis. More extensive spatiotemporal structures were significantly less recognized by students who drew than by students who reflected. The results suggest that drawing might not be an adequate strategy for learning from visuospatially and spatiotemporally complex animations.