[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: Re: Kendall's tau
At 13:56 13/11/03 -0700, Clint Thompson wrote:
I think you are referring to my article (Newson, 2002). A pre-publication
draft of this paper is available on my website at
I have a question that isn't necessarily related to the syntax
of a command but rather, the appropriateness of the statistic.
As I understand Kendall's tau (otherwise known as Kendalls rank
correlation coefficient), it is most appropriate for ordered
contingency table data; that is, the two variables need to
assume some sort of sequential ordering. My query is this: if
one of my variables is ordered (normal-->moderate-->severe) and
the other is continuous (e.g. proportion of total caloric
intake from saturated fat), can I use Kendall's tau to
ascertain some degree of correlation?? Note that my dataset is
small so I am hesitant to employ a parametric test and I want
to know how much the two variables are correlated, not simply
the existence of an association/distribution equality (K-W
test). I understand there is an issue of the SJ/STB that
addresses an issue similar to this, however, I do not have
access to the journal.
which can be accessed either using a browser or using the Stata -net-
command. You can also find, on this site, the -somersd- package, manuals
for the -somersd- and -cendif- programs, and my UK Stata User Meeting
presentation on the subject.
I hope this helps.
Newson R. Parameters behind "nonparametric" statistics: Somers' D,
Kendall's tau 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: