[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Comparing dependent spearman's correlation coefficients
At 18:37 10/06/2004, Nikos Pantazis wrote:
Such a methodology undoubtedly exists, ie its existence is implicit in the
mathematical statistics that I know. However, it is subject to the caution
that the Central Limit Theorem works slowly for the Spearman correlation
coefficient, even under the ideal conditions of a null hypothesis. This was
pointed out by Kendall and Gibbons (1990), who also pointed out that the
Central Limit Theorem works a lot faster for Kendall's tau.
A coleague of mine asked me the following question. Is
there a way to test the equality of two spearman
correlation coefficients e.g H0:r(y,x1)=r(y,x2). We
are aware for the existence of similar methodology for
parametric (Pearson's) correlation coefficients.
The -somersd- package, downloadable from SSC, can be used together with the
-lincom- command of official Stata to calculate confidence intervals for
and also the corresponding differences in Somers' D. For more about how to
do this, type in Stata
net from http://www.kcl-phs.org.uk/rogernewson/
and download my Stata Journal paper on the subject (Newson, 2002), and also
my Stata User Meeting presentation on the subject. See also the .pdf
manuals distributed on SSC, and on my website, with the -somersd- package.
I hope this helps.
Kendall MG, Gibbons DJ. Rank Correlation Methods. 5th ed. New York: Oxford
University Press; 1990.
Newson R. Parameters behind "nonparametric" statistics: Kendall's tau,
Somers' D and median differences. The Stata Journal 2002; 2(1): 45-64.
Lecturer in Medical Statistics
Department of Public Health Sciences
King's College London
5th Floor, Capital House
42 Weston Street
London SE1 3QD
Tel: 020 7848 6648 International +44 20 7848 6648
Fax: 020 7848 6620 International +44 20 7848 6620
or 020 7848 6605 International +44 20 7848 6605
Opinions expressed are those of the author, not the institution.
* For searches and help try: