A Comparison of two Robust Estimation Methods for Business Surveys. (23rd June 2016)

• Record Type: Journal Article
• Authors: Graham Clark, Robert; Kokic, Philip; A. Smith, Paul
• Is Part Of: International statistical review. Volume 85:Number 2(2017)
• Page Start: 270
• Page End: 289
• Publication Date: 2016-06-23
• Subjects: Bootstrap -- mean squared error -- M‐estimation -- movement estimation -- influential values -- outliers -- robustness -- sample survey -- Winsorisation -- Winsorization; Statistics -- Periodicals; Statistics -- Bibliography -- Periodicals; Statistics -- Bibliography; Statistics -- Periodicals; Statistique; Statistique -- Périodiques; Statistique -- Bibliographie -- Périodiques; Statistique -- Étude et enseignement -- Périodiques; Statistique -- Étude et enseignement -- Bibliographie -- Périodiques; Ressource Internet (Descripteur de forme); Périodique électronique (Descripteur de forme); 519.2
• DOI: 10.1111/insr.12177
• Languages: English
• ISSNs: 0306-7734

Abstract

Summary: Two alternative robust estimation methods often employed by National Statistical Institutes in business surveys are two‐sided M‐estimation and one‐sided Winsorisation, which can be regarded as an approximate implementation of one‐sided M‐estimation. We review these methods and evaluate their performance in a simulation of a repeated rotating business survey based on data from the Retail Sales Inquiry conducted by the UK Office for National Statistics. One‐sided and two‐sided M‐estimation are found to have very similar performance, with a slight edge for the former for positive variables. Both methods considerably improve both level and movement estimators. Approaches for setting tuning parameters are evaluated for both methods, and this is a more important issue than the difference between the two approaches. M‐estimation works best when tuning parameters are estimated using historical data but is serviceable even when only live data is available. Confidence interval coverage is much improved by the use of a bootstrap percentile confidence interval.