Premium
Distribution‐free confidence intervals for the standardized median
Author(s) -
Staudte Robert G.
Publication year - 2013
Publication title -
stat
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.61
H-Index - 18
ISSN - 2049-1573
DOI - 10.1002/sta4.29
Subject(s) - quantile , confidence interval , outlier , statistics , sample size determination , variance (accounting) , range (aeronautics) , mathematics , location parameter , cdf based nonparametric confidence interval , scale (ratio) , robust confidence intervals , econometrics , probability distribution , engineering , cartography , geography , accounting , business , aerospace engineering
Assuming that data can be modeled by an unknown location‐scale family of continuous distributions, the aim is to robustly estimate an effect size defined as the median divided by an interquantile range, where the symmetric quantiles are fixed and to be chosen. It is shown that the sample version of this effect size can be variance stabilized, assuming only that the location family density is continuous and positive at the median and the quantiles defining the interquantile range. Confidence intervals for this effect size, which do not require knowledge of the underlying population, are derived and assessed for coverage. These methods are highly resistant to outliers and simple to implement on freely available software. Copyright © 2013 John Wiley & Sons Ltd