The Construction and Analogical Transfer of Symbolic Visualizations

Two studies explored whether adolescents can and will construct abstract visualizations to structure complex information. Experiment 1 showed that students can structure novel information with visualizations. Seventh-, 9th-, and 10th-grade students were directed to construct visualizations of complex transmission problems in biology (e.g., epidemiology). Two-thirds of the resulting visualizations at each grade level captured the structure of the problems. The 9thllOth graders primarily constructed path diagrams (i.e., directed graphs), and the 7th graders constructed more original visualizations. Experiment 2 showed that students will analogically transfer specific visualizations and will transfer the strategy of visualizing. On preand posttests 7th graders solved transmission problems without cues to visualize. During an intervention, students in two treatments constructed and learned visualizations for three problem types. In the Path treatment, one problem type involved path diagrams. In the No-Path treatment an alternate problem type was used as a filler visualization task. Overall, 50% more students used visualizations on the posttest compared to the pretest. Students in the Path treatment analogically transferred the path diagram to the posttest, whereas students in the No-Path treatment spontaneously constructed alternative visualizations. These findings suggest that instruction in visualizing may develop a strategy that students can and will use to understand the structure in complex and novel information. An important aspect of learning new information is to understand the structure among concepts. There are numerous approaches to help students develop a structural understanding of novel information, including the use of analogies (Polya, 1945), concept maps (Schmid & Telaro, 1990), advanced organizers (Ausabel, 1968), planning (Pea & Kurland, 1987), data organization (Underwood & Underwood, 1987), and illustrations (Allender, 1991). These are general methods, because they can be applied to a variety of informational sources; however, there is limited evidence that students can or will spontaneously use general methods without direction (e.g., Detterman, 1993; Pea & Kurland, 1987; Sweller, 1990). In this article I present evidence that another approach for understanding structure-symbolic visualizing-may provide a general-purpose strategy that students can and will use for novel information. Symbolic visualizations-visualizations that do not resemble their referents (Cummins, 1989)-are abstract, nonsentential representations of structure. For example, to convey the cyclical nature of photosynthesis and respiration one could draw a path diagram (i.e., a directed graph). Like drawing an analogy, a good visualization can reveal structure in novelty (Gentner, 1989; Novick, 1990); however, unlike analogies, symbolic visualizations explicitly indicate structure (Breuker, 1984) and are well suited to numerous relations (Bertin, 1983). The utility of