|
Activity Number:
|
244
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #300276 |
|
Title:
|
Space-Time Environmental Interaction Modeling for Small-Area Health Outcomes
|
|
Author(s):
|
Andrew B. Lawson*+ and Ahmed Al Hadrani
|
|
Companies:
|
University of South Carolina and University of South Carolina
|
|
Address:
|
Dept of Epoidemiology & Biostatistics, Columbia, SC, 29208,
|
|
Keywords:
|
spatial ; temporal ; mapping ; disease ; environmental ; exposure
|
|
Abstract:
|
The analysis of space-time data is underdeveloped in general, and restricted to smoothing models for disease risk estimation in particular space-time applications. Convolution models with autoregressive temporal effects are common with interactions modeled as either independent or nonseparable. Often in environmental applications the focus is in estimation of regression parameters relating risk to exposures or their surrogates. Distance and direction functionals are important. However conventional convolution models fair badly when spatial aliasing of effects could be present. In addition, most previous models ignore the possibility of complex interactions between time and exposure effects. In this talk, we will examine a dynamic model for the linkage between putative source locations and outcomes. We also examine the possibility of modeling latent structures in the exposure time.
|
