Journal of Korean Society for Quality Management (품질경영학회지)

Korean Society for Quality Management (한국품질경영학회)
권호 표지

Outward Testing Procedure for the Identification of Multiple Outliers

다수 이상치 인식(認識)을 위한 외향성 검정 절차

  • Yum, Joon-Keun (Dept. of Statistics, Dongguk University) ;
  • Kim, Jong-Woo (Dept. of Mathematics Education, Cheju National University of Education)
  • 염준근 (동국대학교 통계학과) ;
  • 김종우 (제주교육대학교 수학교육학과)
  • Published : 1996.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

This article is concerned with procedures for detecting multiple y outliers in linear regression. The outward-testing procedure, which is controled by the initial subset and the minimum residuals, is suggested by two phases. The performance of this procedure is compared with others by Monte Carlo techniques and found to be superior. The procedure, however, fails in detecting y outliers that are on high-leverage cases in Phase 1. Thus, we proposed ELMS algorithm for a set of suspect observations, in Phase 1. In Phase 2, the proposed testing is conducted using the studentized residuals to see which of the suspect cases are outliers. Several examples are analyzed.

Keywords