A Pseudo-Likelihood Approach for Hierarchical Linear Models Under Complex Designs

• Activity Number: 653
• Type: Contributed
• Date/Time: Thursday, August 7, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Survey Research Methods Section
• Author(s): Junvie Pailden*+
• Companies: Southern Illinois University Edwardsville
• Keywords: pseudo-likelihood ; hierarchical models ; multi-level models ; complex designs ; design weights ; stratified sampling

Abstract

Hierarchical linear models account for different levels of aggregation that may be present in data from survey samples. Inferential procedures that take account of complex survey design features are well established for single-level (or marginal) models. However, the design weights that are required to reflect complex designs are rarely considered under hierarchical linear models. In this paper, we develop a pseudo-likelihood approach procedure to incorporate the design weights for hierarchical linear models model framed in finite population. The proposed pseudo-log-likelihood function is an unbiased estimator of the log-likelihood function when the entire population is sampled. A limited simulation study is presented to compare the proposed approach from models with and without the design weights. Results show significant reduction in the mean squared error using the proposed approach especially when the design weights are related to the response values.