An Examination of Using EWMA Charts for Monitoring Michigan Geospatial Data

• Activity Number: 674
• Type: Contributed
• Date/Time: Thursday, August 4, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Paul Stephenson*+ and Marco Benedetti
• Companies: Grand Valley State University and Grand Valley State University
• Keywords: Moran's I statistic ; EWMA control chart ; biosurveillance ; Spatial surveillance

Abstract

Geographic Information Systems (GIS) are often used to accumulate and present data. Visualization of GIS data through maps is useful in displaying emerging patterns, but they can be misleading. In addition to visualization tools, one needs appropriate statistical methods to determine whether the observed data from a series of maps (constructed over a period of time) indicate that a systematic trend is developing.

Data from Michigan will be the subject of several applications, and the authors will examine how the value of Moran's I statistic changes for patterns of increasing spatial dispersion. The authors will demonstrate how to create an EWMA control chart for global spatial statistics and monitor these charts over time. In the event that a control chart signals, one can then examine the local spatial statistics to determine the source of the signal.

In an effort to examine the power associated with these monitoring procedures, the authors will also simulate the effect of an increase in one, two of three regions in the state of Michigan. The results of these simulations will be used to characterize the effects of shifts in the spatial dispersion across the state.