Diagnostics for comparing robust and least squares fits

Joseph W. McKean, Joshua D. Naranjo, Simon J. Sheather

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


Is a simple least squares (LS) fit appropriate for the data at hand? How different would a more robust estimate be from LS? Is a high breakdown estimator necessary, or is a highly efficient robust estimator sufficient? We propose diagnostics which help answer these questions by measuring the difference in fits between least squares and, successively, a highly efficient robust estimate and a bounded influence robust estimate. Our diagnostic TDBETAS measures the overall change in parameter estimates among these three fits, while the casewise diagnostic CFITS measures change in individual fitted values. We also propose a plot based on CFITS which provides an effective graphical summary of underlying data structure.

Original languageEnglish
Pages (from-to)161-188
Number of pages28
JournalJournal of Nonparametric Statistics
Issue number1-3
StatePublished - 1999


  • Bounded influence
  • GR-estimates
  • High breakdown
  • Linear model
  • Outlier
  • R-estimates
  • Rank based methods
  • Regression diagnostics

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty


Dive into the research topics of 'Diagnostics for comparing robust and least squares fits'. Together they form a unique fingerprint.

Cite this