COMPSTAT 2018: Start Registration
View Submission - COMPSTAT2018
A0231
Title: Spatial small area smoothing models for handling survey data with nonresponse Authors:  Kevin Watjou - University Hasselt (Belgium) [presenting]
Christel Faes - Hasselt University (Belgium)
Andrew Lawson - Medical University of South Carolina (United States)
Rachel Carroll - NIH (United States)
Mehreteab Aregay - Novartis (United States)
Yannick Vandendijck - Hasselt University (Belgium)
Russell Kirby - University of South Florida (United States)
Abstract: Spatial smoothing models play an important role in the field of small area estimation. In the context of complex survey designs, the use of design weights is indispensable in the estimation process. Recently, efforts have been made in the development of spatial smoothing models, in order to obtain reliable estimates of the spatial trend. However, the concept of missing data remains a prevalent problem in the context of spatial trend estimation as estimates are potentially subject to bias. We focus on spatial health surveys where the available information consists of a binary response and its associated design weight. Furthermore, we investigate the impact of nonresponse as missing data on a range of spatial models for different missingness mechanisms and different degrees of nonresponse by means of an extensive simulation study. The computations were done in R, using INLA and other existing packages. The results show that weight adjustment to correct for nonresponse has a beneficial effect on the bias in the missing at random (MAR) setting for all models. Furthermore, we estimate the geographical distribution of perceived health at the district level, based on the 2011 Belgian Health Interview Survey.