5163.0: Wednesday, November 15, 2000 - Board 2

Abstract #15770

Graphically exploring atlas of cancer mortality

Edward C. Chao, Research, MathSoft, Inc, 1700 Westlake Avenue North, Suite 500, Seattle, WA 98109, (206) 283-8802 ext. 227, echao@statsci.com

To study health outcomes and environmental factors simultaneously, methods for exploratory data analysis of data with both space and time variables are very useful. A graphical environment is presented. The system provides the capability of dynamic graphics, in which a graphical window using the ArcView map and another graphical window for displaying statistics graphics are dynamically linked together. Such an environment is powerful for spatial-temporal data. An application to the atlas of cancer mortality is demonstrated. Traditionally, either age-specific mortality rates or cancer mortality rates by regions are displayed individually. Mixed effects models are applicable in evaluating geographic and temporal variations. It requires at least two graphical windows to visualize the impact of correlation structures. For residuals analysis, it is also of interest to see the regions of those outliers in the residuals plot. Such exploratory methods are presented to show the cancer mortality rates by ages and regions.

Learning Objectives: Explore and visualize data with space and time variables simultaneously

Keywords: Cancer, Statistics

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: ArcView of ESRI, Splus of MathSoft
I have a significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.
Relationship: Employment with MathSoft, Inc.

The 128th Annual Meeting of APHA