Dealing with Outliers: Robust, Resistant Regression

Leslie Glasser
Nanochemistry Research Institute, Department of Applied Chemistry, Curtin University of Technology, Perth WA 6845, Australia
J. Chem. Educ., 2007, 84 (3), p 533
DOI: 10.1021/ed084p533
Publication Date (Web): March 1, 2007
History, Education, and Documentation

Abstract

Least-squares linear regression is the best of statistics and it is the worst of statistics. The reasons for this paradoxical claim, arising from possible inapplicability of the method and the excessive influence of "outliers", are discussed and substitute regression methods based on median selection, which is both robust and resistant, are introduced. A simple Excel macro to perform median-based linear regression, as suggested by Theil and by Siegel, is described and made available.

Keywords (Audience):

Graduate Education / Research

Keywords (Feature):

Research: Science and Education

Keywords (Pedagogy):

Textbooks / Reference Books

Keywords (Subject):

Calcium
View: Hi-Res PDF | PDF w/ Links

Tools

SciFinder Links

SciFinder subscribers: Click to sign in | Not a SciFinder subscriber? Learn more at www.cas.org

Explore by:

History

  • Received: August 03, 2009

Recommend & Share

  • Share on ACS NetworkACS Network
  • Add to FacebookFacebook
  • Tweet ThisTweet This
  • Add to CiteULikeCiteULike
  • Add to NewsvineNewsvine
  • Digg ThisDigg This
  • Add to DeliciousDelicious

Related Content