Dear JP,
according to the help menu, you can only specify a second or third matchvar if
you perform a nearest-neighbour, or kernel-based, matching. The thing is, I
want to perform a "several nearest-neighbour" matching and specify a second
matchvar.
Is there some way I can get around this problem?
//pathric
Joao Pedro W. de Azevedo (2003-01-17 14:07):
Dear Pathric,
PSMATCH allows you to match on up to three variables.
Read the option "on( )" on the psmatch help menu.
You can save the predicted value of your probit or logit, and then use it as
one of the three matching criterias with the option "on". Please not that if
a second or third matchvar is specified, psmatch will look for the closest
match in terms of the Mahalanobis distance constructed from the two or three
variables.
All the best,
JP/
-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of
[email protected]
Sent: Friday, January 17, 2003 10:01 AM
To: [email protected]
Subject: st: psmatch-command
Hello,
the psmatch-command does not allow me to match on more than one variable
performing a several nearest neighbour matching.
I need to match on two variables.
Does anybody know how to do this?
Thanks in advance.
//pathric h�gglund
Schonlau, Matthias (2003-01-16 21:27):
I am not sure whether the problem is clear. Propensity scoring and nearest
neighbour
are two very different techniques. I am also unclear as to what stata
function this problem refers to.
Matt
-----Original Message-----
From: [email protected] [mailto:[email protected]]
Sent: Thursday, January 16, 2003 7:11 AM
To: [email protected]
Subject: st: Propensity score matching
Hi!
I want to perform a several-nearest-neighbour matching and match on an
additional (dummy) variable besides "pscore".
To just include this variable in the psmatch statement is not possible. My
solution to this problem is to divide the sample and perform separate
matchings,
and then "append" the two parts before I compare outcome means.
If this is right I would receive similar results as if I actually could
include
this second variable in the psmatch statement. However, testing the
correspondance between these procedures in a nearest neighbour matching, the
results differ a little bit. When I perform a Kernal matching, the results
differ a lot.
The question is why these approaches do not correspond.
Since the number of observations in the two approaches does not fully
correspond I
suspect there is a common support issue.
Do you have an idea how to solve this problem?
Greatfully!!
/pathric
*
* 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/
*
* 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/
*
* 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/
*
* 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/
*
* 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/
