Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: Correlations for censored data


From   "Roger B. Newson" <r.newson@imperial.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Correlations for censored data
Date   Wed, 23 Oct 2013 12:52:48 +0100

I don't know about -tetrachoric-. However, in rank statistics, the -somersd- SSC command (with the -taua- option) can produce Kendall tau-a rank correlations for left-censored and right-censored data (using the -cenind()- option).

I don't know if it is a good idea to enter the matrix of Kendall tau-a stats into a factor analysis method, though. FWIW, when I was at Reading for my MSc, I was told that the Department officially disapproved of factor analysis.

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 23/10/2013 12:29, Seed, Paul wrote:
Dear Statalist,
I have data for a group of subjects on a large number of biomarkers that are sometimes measured,
sometimes only recorded as "below the limit of detection", and sometimes even "above the limit of accuracy".

Apart from the censoring, I anticipate that the values will be Normally distributed after log transformation.
So the (transformed) data is censored multivariate Normal, with some underlying distribution _N_(_Mu_, _Sigma_),
where _Mu_ is a vector of means, and _Sigma_ is a matrix of covariances.


Examples:

Subject		Marker1	Marker2	Marker 3
1		<12		20		37
2		144		< 5		28
3		>3000		44		87
4		.		.		.
5		.		.		.

I want to reduce the number of biomarkers via factor analysis.
Is it possible to estimate the true (Pearson's product moment) correlation
between each pair of biomarkers (i.e. what I would get if I had the actual values).
I am hoping for something like the -tetrachoric-  command; or at least some
advice about how to handle the maximum likelihood calculations.

Paul T Seed, Senior Lecturer in Medical Statistics,
Division of Women's Health, King's College London
Women's Health Academic Centre, King's Health Partners
(+44) (0) 20 7188 3642.


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


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index