Realism as a retrieval cue: Evidence for concreteness‐specific effects of realistic, schematic, and verbal components of visualizations on learning and testing

• Published in: Human behavior and emerging technologies Vol. 3; no. 2; pp. 283 - 295
• Main Authors: Skulmowski, Alexander, Rey, Günter Daniel
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.04.2021
• Subjects: Between-subjects design; concreteness; Cues; Experiments; e‐learning; human cognition; human‐computer interaction; instructional design; Learning; Multiple choice; Realism; Retention; Retrieval; retrieval cues; retrieval specificity; Technology; Virtual reality; Visualization; visualizations
• ISSN: 2578-1863; 2578-1863
• DOI: 10.1002/hbe2.209

Previous research in technology‐enhanced learning suggests that realism in visualizations can overwhelm learners and lower their learning performance. However, some studies offer evidence that realism may be helpful in learning tests that require detailed visual knowledge. A potential mechanism may be that realistic details act as retrieval cues (i.e., visual cues that facilitate retrieval during testing). In order to assess the impact of realism in learning and testing, we conducted an experiment (N = 40) using realistic and schematic renderings generated for this study. Realism was varied as a between‐subjects factor while participants completed both a schematic and realistic retention test. As hypothesized, a realistic test visualization primarily benefits those who had learned using a realistic rather than a schematic visualization. In addition, the realistic test led to higher retention scores regardless of the realism level of the learning materials. A second experiment (n = 50) was conducted to assess whether schematic visualizations can be made more learner‐friendly by including verbal descriptions that enhance the less detailed visualizations with additional cues. The experiment revealed that learning with schematic visualizations can be fostered using verbal descriptions for a multiple‐choice retention test, but at the cost of lower scores in an image‐based retention test. Both studies suggest that matching the realism and concreteness in the learning and testing stages lead to the best performance. Our results have implications for the design of a range of technology‐enhanced learning contexts such as virtual reality education.