I did not get feedback by mail but I got a good reference from Zahaf Toufik, one
of my collegues, which is worth sharing:
look at http://stat.tamu.edu/~aolmsted/paper.html.
Brigitte.
From: Brigitte B Cheuvart on 04/09/2000 16:02
To: [log in to unmask]
cc:
Subject: Hodges-Lehmann estimator
I have been advised to use the Hodges-Lehmann estimator to evaluate the shift
between 2 distributions. I have found that this estimator can be used for
evaluating E50 (ie kind of median for a distribution). Has anyone ever used
that estimator to test/quantify the shift between 2 independent distributions?
Thks in advance for your help.
Brigitte.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|