Search Results

Source: Jossey-Bass
Resulting in 1 citation.
1. Wu, Lawrence L.
Robust M-Estimation of Location and Regression
In: Sociological Methodology. N.B. Tuma, ed. San Francisco, CA: Jossey-Bass, 1985
Cohort(s): NLSY79
Publisher: Jossey-Bass
Keyword(s): Research Methodology; Wages

Permission to reprint the abstract has not been received from the publisher.

This chapter discusses one class of robust estimators, the M-estimators, which are designed to perform well even when the observed data deviate from the common assumption of a Gaussian (or normal) distribution. An advantage of M-estimators is that they permit the estimation of both central tendency in univariate problems and regression coefficients in multivariate problems. Several empirical examples are given, including an analysis of data taken from the 1979 NLSY on hourly wages of a sample of 19-year-old white males. Differences between OLS and M-estimates of regression coefficients appear due to one extreme outlier and several other cases with low reported wages.
Bibliography Citation
Wu, Lawrence L. "Robust M-Estimation of Location and Regression" In: Sociological Methodology. N.B. Tuma, ed. San Francisco, CA: Jossey-Bass, 1985