[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Stratified categorical matched analysis

From   "Svend Juul" <SJ@SOCI.AU.DK>
To   <>
Subject   Re: st: Stratified categorical matched analysis
Date   Sat, 27 Oct 2007 10:49:12 +0200

Diego wrote
I have to perform a test for categorical variables stratified 
by the groups of patients I have (two groups). 
I thought to use McNemar for that, but to perform a stratified 
categorical analysis I should use Mantel-Haenzel (is that 
correct) ? 
I tried to use McNemar (mcc command) with the "by" option but 
it does not accept it.
So the question : how could I perform a stratified McNemar test
for paired categorical data ? 
I may misunderstand your description, but if this is correct:
1. The outcome is dichotomous, like dead/alive
2. You have matched pairs of patients, one exposed to A, the 
   other to B
3. Your patient-pairs fall in two groups, e.g. living in two
   different areas og suffering two different diseases (X and Y).
4. The two patients in a matched pair come from the same group
   (X or Y).
then you can make a standard matched analysis (-mcc-) because
due to the matching principle the two groups (X and Y) are not
mixed up.
You may, however, be interested in any differences in effect
between group X and Y. In that case you could make separate 
-mcc- analyses for the two groups. Another option is conditional
logistic regression (-clogit-).
Hope this helps
Svend Juul
Institut for Folkesundhed, Afdeling for Epidemiologi
(Institute of Public Health, Department of Epidemiology)
Vennelyst Boulevard 6 
DK-8000 Aarhus C,  Denmark 
Phone, work:  +45 8942 6090 
Phone, home:  +45 8693 7796 
Fax:          +45 8613 1580 

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index