Article:

Classifying instructional visualizations: A psychological approach

The use of instructional visualizations has become very popular in the last decades, especially due to the rapid development of technical solutions. Visualization is a broad concept with diverse dimensions. Visualizations do not only vary with regard to their structural features (e. g., dynamism, interactivity), but they also have different functional features (e. g., decoration, representation, organization). Psychological research has investigated these two dimensions particularly in learning contexts. Moreover, a third dimension of visualizations, namely the depicted contents, has been identified. However, the high variability of visualizations is challenging with regard to generalizability of empirical research results found with a specific type of instructional visualization. Therefore, a generic classification system is needed to be able to structure the wide body of research. Although, some attempts to address the classification of visualizations have already been made (e. g., Lohse et al. 1994; Rank in 1990), those focus only on either structural or functional dimensions of visualizations. The current work reviews psychological literature on all three aforementioned dimensions aiming at developing a classification system covering the structural features, the functions, and the depicted contents. The classification system allows assigning visualizations to classes of similar visualizations with regard to all three identified dimensions. This classification system (in form of a questionnaire) was evaluated with ten subjects classifying six different visualizations in order to test the interrater reliability, to assess the usability during filling in the questionnaire, and to investigate its applicability to different types of visualizations. Data analyses revealed high, or close to high interrater reliabilities for all six visualizations. These results indicate that a questionnaire-based classification system can be used to objectively classify visualizations.