Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: An urgent request for help :-)


From   Roger Newson <roger.newson@kcl.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: An urgent request for help :-)
Date   Thu, 17 Nov 2005 22:12:56 +0000

You don't state whether these 2 regressions are estimated on the same or different sets of data. If they are estimated on the same set of data, then any solution will probably use -suest- to create a combined covariance matrix for the 2 regressions. If they are from independent sets of data, then the 2 covariance matrices (extracted from the estimation results) can be combined diagonally to produce a combined covariance matrix. Once this combined covariance matrix is produced, you can produce a row vector of the derivatives of

G=(B01-B02)/(B11-B12)

with respect to B01, B11, B02 and B12, and use this vector, and the combined covariance matrix, to compute a covariance for G.

I don't know how much you know about -suest-, Stata matrices, and extracting estimation results. However, -whelp matrix- will introduce you to Stata matrices, -whelp suest- will introduce you to -suest-, and -whelp estimates- will introduce you to estimation results.

Best wishes

Roger


At 20:47 17/11/2005, you wrote:

Hello,
This may be a slightly basic question, but I would really appreciate
a solution... I have two regression estimations which I run
sequentially, and need to get a 95% confidence interval using
coefficients from both these estimations in a non-linear combination.
Specifically, I want to get the (delta-method estimated) 95% conf.
interval for the following expression:

(B01 - B02)/(B11-B12),

 where B01 is the intercept of the 1st regression, B02 is the
intercept of the second equation, B11 is a predictor coefficient of
the 1st equation, and B12 is a predictor coefficient of the 2nd
equation.

The models I'm running are univariate regressions, but I'd appreciate
a multi-variate generalization as well. My question centres around
how one can store and recall estimations from previous regressions
while trying to work with a non-linear combination of their
coefficients.

Thank you, and looking forward to hearing from you,

Sincerely,

Manasi
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

--
Roger Newson
Lecturer in Medical Statistics
Department of Public Health Sciences
Division of Asthma, Allergy and Lung Biology
King's College London

5th Floor, Capital House
42 Weston Street
London SE1 3QD
United Kingdom

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
Email: roger.newson@kcl.ac.uk
Website: http://phs.kcl.ac.uk/rogernewson/

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

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   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   |   What's new   |   Site index