When random effects are correlated with survey sample design variables, the usual approach of employing individual survey weights (constructed to be inversely proportional to the unit survey inclusion probabilities) to form a pseudo-likelihood no longer produces asymptotically unbiased inference. We construct a weight-exponentiated formulation for the random effects distribution that achieves approximately unbiased inference for generating hyperparameters of the random effects. We contrast our approach with frequentist methods that rely on numerical integration to reveal that the pseudo Bayesian method achieves both unbiased estimation with respect to the sampling design distribution and consistency with respect to the population generating distribution. Our simulations and real data example for a survey of business establishments demonstrate the utility of our approach across different modeling formulations and sampling designs. This work serves as a capstone for recent developmental efforts that combine traditional survey estimation approaches with the Bayesian modeling paradigm and provides a bridge across the two rich but disparate sub-fields.
Pseudo Bayesian mixed models under informative sampling
Savitsky, T. D., & Williams, M. R. (2022). Pseudo Bayesian mixed models under informative sampling. Journal of Official Statistics, 38(3), 901-928. https://doi.org/10.2478/jos-2022-0039
To contact an RTI author, request a report, or for additional information about publications by our experts, send us your request.
Spatiotemporal trends in three smoothed overdose death rates in US counties, 2012-2020
Improving text classification with Boolean retrieval for rare categories
COVID-19 diagnosis and SARS-CoV-2 strain identification by a rapid, multiplexed, point-of-care antibody microarray