Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: "dependent correlation" or c-statistics


From   Roger Newson <r.newson@imperial.ac.uk>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: "dependent correlation" or c-statistics
Date   Wed, 18 Aug 2010 14:04:43 +0100

It is probably not a good idea to choose a threshold according to whether using that threshold gets a P-value of 0.05 or less. And it might not be a good idea to use a threshold at all, as doing this implies that information about trend within the 2 binary values is not used.

To measure ordinal trend, without converting -ves- to binary, you might like to use the -somersd- package, downloadable from SSC (or from my website if you use Stata 9). If you install this package, then you can type

somersd frailty ves, transf(z) tdist

and you will compute a confidence interval for Somers' D of -ves- with respect to -frailty-. And you can type

somersd ves frailty, transf(z) tdist

and compute a confidence interval for the Somers' D of -frailty- with respect to -ves-.

Alternatively, you can use the Harrell's c parameterization. Harrell's c is linked to Somers' D by the relation D=2c-1, where D is Somers' D and c is Harrell's c. To calculate a confidence interval and P-value for Harrell's c of -ves- with respect to -frailty-, type


somersd frailty ves, transf(c) tdist

and to calculate a confidence interval and a P-value for Harrell's c of -frailty- with respect to -ves-, type

somersd ves frailty, transf(c) tdist

I hope this helps.

Best wishes

Roger


Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: r.newson@imperial.ac.uk
Web page: http://www.imperial.ac.uk/nhli/r.newson/
Departmental Web page:
http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/popgenetics/reph/

Opinions expressed are those of the author, not of the institution.

On 18/08/2010 12:34, Melaku Fekadu wrote:
hello statalisters,

i have two variables: one called 'ves' with values between 1 to 10,
and the other is dichotomus called 'frality' with value either 0 or 1.

i create a categorical variable from 'ves' 0-2 and 3-10 and make a
chi-square test to test a relation with the variable frality, and i
get p>0.05. But when i create another categorical variable of 'ves'
with 0-4 and 5-10 and do the chi-square test i find that p<0.05.

my question: what can i do to get a threshold value of 'ves' where
p-value will be 0.05? somebody told me about a procedure calles
"dependent correlation" or c-statistics in stata.

any help will be appreciated.

thanks
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index