The effects of visualizations and spatial ability on learning from static multimedia instructions

Abstract

Successful learning about physical systems is thought to depend on the development of a mental representation of the system's dynamic behavior, which constitutes a mental model, rather than only its static structure (e.g., Schnotz, 2005). Because dynamic mental models must be generated by learners from static diagrams, learning might be promoted by encouraging learners to visualize motion in static diagrams. However, mental models represent dynamic spatial information that might be difficult to construct for learners with lower spatial ability; they might benefit from instructional designs that support spatial reasoning, such as phase diagrams and depictive arrows. In Experiment 1, participants learned about air pumps, carburetors, and toilet tanks from single phase diagrams, multiphase diagrams, or multiphase diagrams followed by a prediction activity in which they predicted system behavior in novel situations. This prediction activity was expected to implicitly prompt mental visualization of motion. Learning in the latter condition (i.e., with the prediction activity) was significantly better than learning in the single phase condition. In the prediction condition, the enhancing effect of spatial ability on learning outcome was partially mediated by performance in the prediction activity. The mediation suggested that high spatial ability helped participants to accurately visualize the systems as they made predictions, which contributed to better performance on the learning assessment. Experiment 1 assessed visualizations during the prediction activity, whereas Experiment 2 assessed visualizations during the lessons. In two conditions in Experiment 2, participants were explicitly prompted to visualize motion in the system while viewing the lessons. Because learners with lower spatial ability were expected to have difficulty visualizing motion, arrows depicting motion were added in one condition. A baseline condition excluded the arrows and the prompt to visualize motion. In all three conditions, participants viewed multiphase diagrams followed by the prediction activity. Learning outcomes among the three conditions did not differ significantly: Depictive arrows and prompts to visualize motion were not significantly effective. Also, spatial ability did not interact with instructional condition. However, both spatial ability and subjective ratings of attempts to visualize motion were predictive of learning outcome. Overall, results from the two experiments indicated that participants with higher spatial ability were better able than participants with lower spatial ability to generate dynamic mental models from static instructions, particularly when they were implicitly prompted to reason about the system as they made predictions. Learners with lower spatial ability might need other forms of assistance for mental model generation, such as animated instructions.