On Deletion Residuals

Abstract

Residuals based on case deletion has been drawn a great deal of attention in recent years. The prime objective of these residuals is the identification of outliers in linear regression. In this paper we investigate their performances as substitutes of the true randQm disturbances and compare them with the traditionally used least squares residuals. Some distributional properties like moments, coefficients of skewness and kurtosis of deletion residuals are studied and their usefulness in tests for Normality is investigated.