Shape distinctness and segmentation benefit learning from realistic visualizations, while dimensionality and perspective play a minor role

Abstract

Research on the design of visualizations has revealed that realistic visualizations can be beneficial for retention and transfer performance despite their tendency towards causing cognitive load. However, it still can be hard to predict whether a more detailed visualization will be more effective for learning than a more abstract one. The purpose of the reported studies is to investigate whether an emphasis on the specific benefits of realism can enhance learning from realistic visualizations. In three experiments, the dimensionality (i.e., whether flat cross-sections or shapes conveying depth and space are presented), perspective, shape distinctness, and segmentation by color are investigated. In Experiment 1, a visualization of a flat section of the kidney elicited a greater cognitive load if this section was presented as a realistic rendering rather than a schematic drawing, while a more dimensional model of the kidney appears to be harder to cognitively process if it is presented as a drawing. Experiment 2 examined whether a perspective emphasizing depth is more compatible with a realistic rather than a schematic visualization, but this effect failed to reach significance. Lastly, Experiment 3 demonstrated that shape distinctness and segmentation using colors enhance retention performance. The main result of this series of studies is that even minor changes in the presentation of realistic visualizations can have an impact on cognitive load. In conclusion, in order to optimize learning, realistic visualizations should make use of depth, emphasize semantic information by increasing the distinctness of shapes, and use segmentation using colors or different surface materials.