Variance Estimation for High-Income Tables

• Activity Number: 139
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Wei Qian*+
• Companies: Statistics Canada
• Keywords: high income tables ; Longitudinal Administrative Databank ; percentile group ; Taylor linearization

Abstract

Statistics Canada produces high income statistics that provide information on demography, taxation, and income for high income populations. These statistics involve estimation of percentiles and quantities in percentile groups using a sample that covers around 1/5 of the Canadian population. For variance estimation, re-sampling methods may take unacceptable time due to the extremely large sample size; linearization cannot be applied directly because of the non-smoothness of the estimators. In this paper we propose a weighted estimating equations approach to derive linear variance estimators. Data from Prince Edward Island (P.E.I) is used to illustrate the results obtained.