Showing papers by "Joseph W. McKean published in 1987"
Journal Article•
TL;DR: In this article, an estimator for the scale parameter is proposed, which does not require the symmetry of the error distribution, and a Monte Carlo study is presented to examine the performance of the estimator in a small sample setting.
Abstract: In the analysis of a general linear model a scale parameter arises in the standard deviations of rank estimates and as a standardizing constant for rank test statistics. An estimator is proposed for this scale parameter which does not require the symmetry of the error distribution. A proof of its consistency is given. A Monte Carlo study is presented to examine the performance of the estimator in a small sample setting.
102 citations
TL;DR: In this paper, the least-absolute deviation or l 1 analysis of a linear model is an important alternative to the classical least squares analysis from the point of view of efficiency for longer-tailed error distributions and robustness to the presence of outliers.
44 citations
TL;DR: In this article, the authors used the sign statistic to test the median of the distribution-free confidence intervals and studentized the median by an estimate of its standard error. But they used the standard error as a measure of the confidence of the median.
39 citations