Multivariate Representation Methods for Time Series Geo-referenced Health Statistics Data, MacEachren (PI), Edsall (Co-PI), Sponsor and Collaborating Agency, National Cancer Institute

Geographic Information Systems (GISystems) and mapping can be used for visualizing health statistics "to understand the spatially varying factors that lead to mortality and disease and the variation in those factors for different at-risk groups in the population" (MacEachren 1998, p. 10). The interactive system developed and tested in that report and that developed in Plaisant (1993) both emphasize the need for visual representations and dynamic queries in spatiotemporal health data analysis. The initial stage of this research extends the conceptual frameworks and results of those studies by developing methods for addressing tasks requiring the assimilation of multiple variables, observations, and cases at once, through multiple, dynamically linked, representation forms.

The project explores, specifically, the potential of dynamic parallel coordinate plots (PCPs) depicting multivariate attribute data linked to maps depicting the geographic aspects of those data and data-rich active legends that depict the data's temporal component. Dynamic parallel coordinate plots have been demonstrated to be a powerful tool for visual multivariate data analysis (Edsall 1999), but have not been applied to health statistics. The plots employ a novel methodology to visualize more than three dimensions by representing each observation not as a point in a scatter plot but as a series of unbroken line segments connecting parallel axes, each of which represents a different variable. By representing variables as parallel - as opposed to orthogonal - axes, the representation breaks the bonds of two or three-dimensional representations such as scatter plots (Inselberg 1985; Wegman 1990).

The first step in this research (supported directly through a contract from NCI) has been integration of PCPs into our ArcView-based HealthVis environment (see: http://wwww.geovista.psu.edu/NCHS/health.htm) and a formal usability test of that environment for a series of typical data analysis tasks. Results of this experiment will guide work within the dgQG project to both extend HealthVis and (in parallel) develop a web-accessible, Java-based environment that links maps with PCPs.

Edsall, R. (1999). "Development of Interactive Tools for the Exploration of Large Geographic Databases". Proceedings of the 19th International Cartographic Conference, Ottawa, University of Victoria.34B www.geog.psu.edu/~edsall/JSM99/paper.htm

Inselberg, A. (1985). "The Plane with Parallel Coordinates." The Visual Computer 1: 69-97.

MacEachren, A. M. (1998). Design and Evaluation of a Computerized Dynamic Mapping System Interface. Washington, DC, National Center for Health Statistics.

Plaisant, C. (1993). Facilitating data exploration: Dynamic queries on a health statistics map. Proceedings of the The Annual Meeting of the American Statistical Association, Government Statistics Section, San Francisco, CA.18-23

Wegman, J. J. (1990). "Hyperdimensional Data Analysis Using Parallel Coordinates." Journal of the American Statistical Association 85 (411): 664-675.