|
Activity Number:
|
464
- SPEED: Infectious Diseases, Spatial Modeling and Environmental Exposures, Speed 1
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, July 31, 2019 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
|
Abstract #302926
|
|
|
Title:
|
Assessing the Association Between Sex Ratio and Dowry Deaths in Uttar Pradesh Using Spatio-Temporal Random Effects Models
|
|
Author(s):
|
Tomas Goicoa* and MARIA DOLORES UGARTE and Aritz Adin and JIM HODGES
|
|
Companies:
|
Public University of Navarre and PUBLIC UNIVERSITY OF NAVARRE and Public University of Navarre and UNIVERSITY OF MINNESOTA
|
|
Keywords:
|
contraints;
dowry deaths;
INLA;
restricted regression;
spatio-temporal confounding;
women violence
|
|
Abstract:
|
Models with spatially and temporally correlated random effects are useful tools in epidemiology to discover spatio-temporal patterns of risk and to assess associations with potential risks factors. However, when modelling covariates and random effects together, spatial and/or spatio-temporal confounding can affect the estimated association between potential risk factors and outcomes and also inflate the associated standard errors. In this work we consider two possible ways to alleviate this problem. The first one uses constraints on the random effects to avoid identifiability and confounding issues, and the second one is based on a reparameterization of the original model where the random effects are restricted to subspaces orthogonal to the fixed effects. The two approaches give noticeably different fits and associated DIC values. Our final objective is to analyze spatio-temporal patterns of dowry death incidence in Uttar Pradesh, the most populated state in India and to evaluate potential risk factors related to this form of violence against women. Here we discover a negative association between sex ratio and dowry deaths.
|
Authors who are presenting talks have a * after their name.
