| Sumario: | Exploratory data analysis (EDA)—a data-centered, inductive approach to statistical analysis—offers effective instruments for evaluating the quality and integrity of GIS attribute data. In this study, examples demonstrating EDA distribution analyses, correlational statistics, and proximity analysis are provided. An integrated modular software prototype system to operationalize these techniques is described that integrates the mapping and display capabilities of PC ArcView with the statistical capabilities of STATA in an MS-Windows multi-tasking, multiple-windowed environment. The combination of data views that the examples offer is shown to be particularly effective in highlighting distributional extremes, correlational outliers, and spatial anomalies.
|
|---|