Linear regression diagnostics in cluster samples

Abstract

An extensive set of diagnostics for linear regression models has been developed to handle nonsurvey data. The models and the sampling plans used for finite populations often entail stratification, clustering, and survey weights, which renders many of the standard diagnostics inappropriate. In this article we adapt some influence diagnostics that have been formulated for ordinary or weighted least squares for use with stratified, clustered survey data. The statistics considered here include DFBETAS, DFFITS, and Cook’s D. The differences in the performance of ordinary least squares and survey-weighted diagnostics are compared using complex survey data where the values of weights, response variables, and covariates vary substantially.