Scientific Visualization as an Interpretive and Expressive Medium Douglas N. Gordin Daniel C. Edelson Louis M. Gomez Northwestern University, School of Education and Social Policy, 21 15 N. Campus Drive, Evanston IL, 60208 Introduction As a powerful technique for the visual representation of complex data, scientific visualization offers the potential to help secondary school science students learn through active inquiry. Over a period of several years, we have been conducting research in the design of scientific visualization environments that support inquiry-based learning. The goal is to support students in a multi-stage learning process that culminates in open-ended research projects. In the early stages, learners should gain familiarity with scientific visualization techniques and the phenomena depicted. In the later stages, students conduct investigations that use scientific visualization techniques to explore open-ended questions. Through the process of conducting inquiry with scientific visualization students can gain an understanding of its usefulness for originating and answering scientific questions. The use of scientific visualization, like other media, can be divided into two categories: interpretive and expressive. In interpretive use, the user is primarily a viewer who is attempting to extract meaning from visualizations. In expressive use, the user is an author who is attempting to convey meaning through the construction of visualizations. In the course of this research we have come to recognize the importance of both interpretive and expressive activities for learners. This paper traces our design of software and activities as we have shifted from a primary emphasis on the interpretation of visualizations to one that integrates interpretation and exwession. Our initial software enabled students to view visualizations and our activities called for students to compare and contrast these images. Following a set of experiments in which we observed students drawing visualizations by hand, we have developed new software and activity designs that address some of the limitations of the initial design by incorporating expressive activities and a wider range of data. The more recent software design enables students to create data by "p&ting" visualizations, communicate a point by customizing the appearance of visualizations, and explore hypothetical scenarios by both creating and interpreting visualizations. The setting for these activities is an investigation into global warming. The controversial nature of this subject can serve to enhance student motivation and demonstrate the dynamic quality of scientific knowledge [Linn & Songer 19931. In addition, investigations of global warming can be~substan6ally~aided by scientific visualization, since these investigations requires analyzing large detailed data sets. In the remainder of this paper, we present some experiences with design and use and explore their implications. First, we discuss our observations of the interpretive and expressive uses of visualization by students who were engaged in a unit on global warming. Second, we present three additional design goals based on these experiences. Finally, we describe our most recent design that attempts to achieve these goals. The observations reported in this paper were made in classes that were engaged in a unit on global warming as part of their participation in the Learning Through Collaborative Visualization (CoVis) Project p e a 19931. Interpretive Use of Visualization Supporting student inquiry in global warming led to the design of the Greenhouse Effect Visualizer (GEV) [Gordin, Edelson, & Pea 19951 and introductory activities. The GEV was designed to give students access to atmospheric data describing the earth's energy cycle. The introductory activities focus on the current state of the atmosphere. These activities ask students to explore climatic processes through the interpretation of visualizations. In this section, we describe the design of the GEV and provide an example of its use.