Bayesian model comparison and model averaging for small-area estimation

TL;DR

This paper introduces a methodology for evaluating and averaging over models in small-area estimation, specifically applied to lung cancer mortality data, to improve robustness of area effect inferences.

Contribution

It presents a general approach for model evaluation and averaging in small-area estimation, addressing the challenge of model choice impact on inference.

Findings

Different models lead to varying conclusions about city mortality rates.

Model averaging provides more robust estimates of area effects.

Reanalysis of Missouri lung cancer data demonstrates the methodology's practical impact.

Abstract

This paper considers small-area estimation with lung cancer mortality data, and discusses the choice of upper-level model for the variation over areas. Inference about the random effects for the areas may depend strongly on the choice of this model, but this choice is not a straightforward matter. We give a general methodology for both evaluating the data evidence for different models and averaging over plausible models to give robust area effect distributions. We reanalyze the data of Tsutakawa [Biometrics 41 (1985) 69--79] on lung cancer mortality rates in Missouri cities, and show the differences in conclusions about the city rates from this methodology.