May 13, 2016, 3:00pm
Location: Department of Mathematics and Statistics, Acadia University, Wolfville, NS
Room: Huggins Science Hall 154
Speaker: Changbao Wu, University of Waterloo
Abstract: There has been considerable theoretical development in recent years on empirical likelihood inference for complex surveys, including notably the work by Wu and Rao (2006) and Berger and Torres (2016) on confidence intervals for finite population parameters. Existing approaches to empirical likelihood inference under the design-based framework are not practically useful since they require the first and second order inclusion probabilities of the survey design, which are typically unavailable to survey data users. In this paper we develop empirical likelihood methods for analyzing public-use survey data that contain only the final adjusted and calibrated weights along with suitable replication weights in addition to measurements on survey variables. Asymptotic distributions of the empirical likelihood ratio statistics are derived for parameters defined through estimating equations. Finite sample performances of the empirical likelihood ratio confidence intervals, with comparisons to methods based on the estimating equation theory, are investigated through simulation studies. The proposed approaches make empirical likelihood a practically useful tool for analyzing complex survey data. A few related issues, including missing data problems, are briefly discussed.
The talk is based on joint research with J.N.K. Rao of Carleton University and is sponsored by the CANSSI CRT Project “Statistical Inference for Complex Surveys with Missing Observations”.
Back to Seminar Series Schedule.