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

st: Results different between -pwcorr- and -mkcorr- [Stata 9.1]


From   "Herve STOLOWY" <stolowy@hec.fr>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: Results different between -pwcorr- and -mkcorr- [Stata 9.1]
Date   Sat, 19 Nov 2005 09:36:56 +0100

Dear All:

I use -mkcorr- (downloadable from ssc) to obtain correlations in a file which can be easily opened with excel. In the help file, I read the following text: "Please note that it calls the correlate command rather than the pwcorr command.  Therefore, it will only calculate correlations based on observations in which none of the requested variables are missing".

So, I anticipate differing results between -pwcorr- and -mkcorr-, depending on the size of the sample. 

However, I am puzzled by the following situation:

pwcorr absence1  divergence1 common , sig obs, if sample_japp30==1

mkcorr  absence1  divergence1 common if sample_japp30==1, log(correlation_Table3_PanelA) replace label sig cdec(4)

As the sample should be the same (with the restriction if sample_japp30==1), I anticipate to get exactly the same figures. However, if the coefficients are strictly identical, the p-values are slightly different.

How could I explain this difference? I probably miss something concerning -correlate-.

Best regards

Hervé Stolowy





***********************************************************
Professeur/Professor
Coordinateur du Département/Head of Department
HEC Paris
Département Comptabilité Contrôle de gestion / Dept of Accounting and Management Control
1, rue de la Liberation
78351 - Jouy-en-Josas
France
Tel: +33 1 39 67 94 42 - Fax: +33 1 39 67 70 86
stolowy@hec.fr
http://campus.hec.fr/profs/stolowy/perso/home.htm

*
*   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