|
Activity Number:
|
109
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #305761 |
|
Title:
|
Bayesian Modeling for Ordinal Substrate Size Using EPA Stream Data
|
|
Author(s):
|
Megan Dailey Higgs*+ and Jennifer A. Hoeting and Brian Bledsoe
|
|
Companies:
|
Colorado State University and Colorado State University and Colorado State University
|
|
Address:
|
240 N. McKinley Ave., Fort Collins, CO, 80521,
|
|
Keywords:
|
ordinal data ; spatial models ; Bayesian ; categorical data
|
|
Abstract:
|
Substrate size can be indicative of stream health and an important determinant of habitat suitability for fish and macroinvertebrates. The U.S. Environmental Protection Agency (EPA) collected data at 485 stream sites in Oregon and Washington between 1994 and 2004. The measurement and recording protocol for substrate size resulted in ordered categorical data. Previous attempts at building successful predictive models for substrate size have treated the variable as a continuous measurement. We investigate methods to model it as an ordinal categorical variable rather than naively assuming it is continuous. Additionally, we incorporate the spatial correlation inherent in the data using Bayesian methods to build an ordinal categorical spatial model.
|
