|
Activity Number:
|
534
- Contributed Poster Presentations: Section on Statistics in Epidemiology
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2018 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
|
Abstract #329821
|
|
|
Title:
|
Distance-Weighted Predictor Models to Estimate the Spatial Scale of Built Environment Health Effects
|
|
Author(s):
|
Adam Peterson* and Brisa N. Sanchez and Emma V Sanchez-Vaznaugh
|
|
Companies:
|
University of Michigan and University of Michigan and San Francisco State University
|
|
Keywords:
|
Built-Environment;
Spatial Scale;
Stan;
Hierarchical Model;
Food Environment;
Bayesian
|
|
Abstract:
|
Identifying the spatial scale at which the availability of built environment features may confer health effects has been a challenging problem in the built environment literature. We propose a new modeling framework for estimating the spatial scale at which environmental features affect an outcome of interest. The proposed hierarchical model uses distances between study participants' addresses and the locations of environmental features (e.g., grocery stores) to define weighted counts of one or more classes of environmental features. The weights are parametrized as a function of distance according to a function with interpretable parameters; health effects are thus quantified by both the spatial scale and the magnitude of the association. We implement this model in R via the No U-Turn Hamiltonian Monte Carlo Sampler software, Stan. The model is used to examine the spatial scale and magnitude of association between the availability of multiple food environment features near schools and children's weight status.
|
Authors who are presenting talks have a * after their name.
