Visualization in Science Education - PDF document

Visualization in Science Education 129 Using Spatial Data Visualization to Motivate Undergraduate Social Science Students Richard LeGates, San Francisco State University, San Francisco, CA sociology, political science, anthropology, urban studies, andability. Most undergraduate social science majors takethey are dreaded requirements for the major.major. Subjecting idealistic students who hate math to rig-American City,Ó ÒRace, Class, and the Environment,Ó or ÒSocialsocial science discipline as a vehicle to make the world better.NSF ÒSpace, Culture, and Urban PolicyÓ CCLI Educational eeing Social Relations and Social PolicyOptions Through GIS GIS software links computerized maps to attribute tablesthat contain data about map features. In addition to physical features such as rivers and roads, a GIS can displayinformation about social and policy phenomena such as 130 Visualization in Science Education race, income, housing affordability, voting behavior, histor-Until recently, GIS software was expensive and technicallyMicrosoft Word document or Excel spreadsheet. Studentstures distributed in geographical space?Ó, ÒAre patterns dis-output more elegantly, they can produce ever more revealingthemselves tailor. A beginning student who knows onlycities whose 2015 population is expected to be greater thanthe answer visually. Another query with Boolean operatorsto be over 10,000,000 and to be under $10,000 U.S. will instantly produce a new map.scientiÞc inquiry.Santa Barbara, geography professor Keith Clarke, the authorof a leading undergraduate GIS textbook (Clarke, 2002)the author of the Environmental Science Research InstituteÕs Inverting the Role of Exploration andisualization in Social Research Historically, most social scientists research followed a rigidthen testing the hypotheses using the data. Yale political science professor John Tukey (Tukey, 1977) inverted this waythem, Tukey advocated exploratory data analysis. He encour-continuously as they explored data. Tukey also invented ways Figure 1. Cities of the future: world population estimates for 2015. Visualization in Science Education 131 research reports. Contrary to TukeyÕs method, this approachafter the analysis is complete. Regrettably, even today, mostthat was as revolutionary as TukeyÕs theories aboutmeaning to the viewer. Bertin developed a body of theorystanding the data and reÞning the research. Like Tukey, BertinentiÞc data visualization. In addition to Tukey and Bertin,ale political science professor, Edward Tufte, has writtendata (Tufte, 1983, 1990, 1997). William Cleveland, a scientistStuart Card, Jock D. Mackinlay, and Ben Shneiderman haveOur pedagogical model draws upon this body of theory.zation. We go even further, arguing that data visualization Visualization First iments, and secondary data analysis. There is a better way.of these courses. Accordingly, we advocate starting thesebetween poverty, race, and the location of toxic sites; eco-answering these kinds of Òwhat is where,Ó Òhow many,Ó Ònear 132 Visualization in Science Education analysisÑindeed all scientiÞc inquiry.tant dimension of voting behavior. An economist analyzingare not uniformly distributed in metropolitan areas today.scientists can make good use of spatial analysis. Verbal, tab-ular, and statistical representations of where phenomena How to Teach Spatial Analysis and Data amuk, project coÐprincipal investigator, is developing a Globally/Acting Regionally. By Òmodule,Ó we mean 1) a short,visual, paperback book with material appropriate for foursessions of an upper-division undergraduate social scienceresearch methods or data analysis class; 2) GIS data sets toaccompany the modules; and 3) additional resources on awebsite accompanying the modules. Thinking Globally/Acting Regionally all of the social sciences. This Þlls a notable gap. Currently,GIS courses (Bolstad, 2002; Clarke, 2002; DeMers, 2000;Heywood and Cornelius, 2002; Theobald, 2003). There areGIS software is evolving very rapidly, and most GIS textstitled ÒYour TurnÓ at the end that asks students to repeat Visualization in Science Education 133 social reality are very much in order. Conclusion zation, beginning right away, motivates students becauserists such as John Tukey and Jacques Bertin have advocatedinquiry. This NSF CCLI-EMD grant will develop materials that LIOGRAPHY Bertin, J. (translated by William J. Berg). 1967. Reprint 1983. Semiology of Graphics: Diagrams, Networks, Maps. Madison, WI: University of Wisconsin Press.Bolstad, P. 2002. S Fundamentals. White Bear Lake, MN: EiderBossard, E. G. 2005. Envisioning Neighborhoods. edlands, CA:ess, forthcoming.Card, S. K., J. D. Mackinlay, and B. Shneiderman. 1999. in Information Visualization: Using Vision To Think. ancisco, CA: Morgan Kaufmann. Getting Started with Geographic Information 4th ed. New York: Prentice-Hall.Cleveland, W. 1994. The Elements of Graphing Data. tion. Murray Hill, NJ: AT&TBell Labs. undamentals of Geographic Information 2nd ed. New York: Wiley.Heywood, I., S. Cornelius, and S. Carver. 2002. An Introductionto Geographical Information Systems. 2nd ed. New York:ongley, P. A., M. F. Goodchild, D. J. Maguire, and D. W. Rhind. Geographic Information Systems: Principles,echniques, Management and Applications. New York: Wiley. The ESRI Guide to GIS Analysis, Volume 1:Geographic Patterns and Relationships. edlands, CA: ESRINational Center for Geographic Information and Analysis. 2000. Core Curriculum in GIScience Santa Barbara. Santa Barbara,A: National Center for Geographic Information andNyerges, T., and R. Golledge. 1997. Asking Geographic Santa Barbara, CA: National Center forGeographic Information and Analysis. The Visual Display of Quantitative Information. Cheshire, CN: Graphics Press. Envisioning Information. Cheshire, CN:Graphics Press. isual Explanations: Images and Quantities,vidence and Narrative. Cheshire, CN: Graphics Press. S Concepts and ArcGIS Methods. Boulder, CO: Natural Resources Ecology Laboratory,Colorado State University.ukey, J. W. 1977. xploratory Data Analysis. eading, MA:Addison-Wesley.University Consortium for Geographic Information Science. The Straw Report: Model Curriculum. 134 Visualization in Science Education Endnotes 1.CCLI-EMD 0228878.2.Nyerges and Golledge (1997) suggest that geographical analysis can seek toanswer the following: Where is it? Where does it occur? What is there? Why is itnot elsewhere? How much is there at that location? How far does it extendalready? Is there regularity in its distribution? What is the nature of that regu-larity? Where is it in relation to others of the same kind? What else is there spa-in the same place? How has it changed spatially (through time)? Why has itspread or diffused in this particular way?3.Clarke suggests geographical questions that undergraduate students can explorewithout statistics such as size, distribution, pattern, contiguity, neighborhood,scale, and orientation. He notes that beginning students can explore whether dis-tributions of things in space are sparse, uneven, random, regular, uniform, scat-tered, or clustered.4.Mitchells widely used introduction to GIS analysis describes how to map themost and the least, density, whats inside, whats nearby, and change. While theprecise methods Mitchell describes are subtle and statistical training providesmore powerful ways to do this, the basic concepts are easily grasped by under-graduate students without statistical training.