st: RE: RE: pscore and attnd

[Rossella Verzulli <rossella.verzulli@unibo.it>]

Dear Statalist,

I'm using the attnd command to identify the group of matched units defined by the nearest neighbor matching method. The alternative psmatch2 allows to calculate the overall weight given to the matched observations (i.e. the frequency with which each observation is used as a match for nearest neighbor mathing). Does anybody know how to calculate this output from attnd? I would like to do some post-matching analysis with the comparison units weighted by the frequency with which each observation is used as a match.

I would be very grateful for any help with this.

Best regards,
Rossella


________________________________________
Da: Rossella Verzulli
Inviato: lunedì 24 marzo 2014 21.44
A: statalist@hsphsun2.harvard.edu
Oggetto: RE: RE: pscore and attnd

Dear Joe

I've just downloaded the updated ado and help files for pscore, which now allows for identification of matched pairs in nearest neighbour matching.

This was very helpful, many thanks for your advice indeed.

Best regards,
Rossella


________________________________________
Da: owner-statalist@hsphsun2.harvard.edu [owner-statalist@hsphsun2.harvard.edu] per conto di Joe Canner [jcanner1@jhmi.edu]
Inviato: lunedì 24 marzo 2014 17.55
A: statalist@hsphsun2.harvard.edu
Oggetto: st: RE: pscore and attnd

Rossella,

I haven't used -pscore- and -attnd- lately (in favor of -psmatch2-), but according to the -attnd- help:

where match_options are

        matchvar(newvarname) matchdta(newfilename) id(varname)

        In nearest neighbor matching, where not all controls are necessarily used in the
        computation of the ATT, it is often useful to identify those observations which were
        actually used.  The option matchvar(newvarname) together with matchdta(newfilename)
        allows to save a dataset containing the treatment indicator, the estimated propensity
        score (with the variable name pscore or the name specified by the user with the
        option pscore(scorevar)), and a dummy indicating whether an observation was used in
        the computation of the ATT. Furthermore, when a variable which identifies
        observations exists in the dataset, users can specify a variable name for this ID
        variable using the option id(varname) and this is saved as well.  In detail, the
        options are described below.

Regards,
Joe Canner
Johns Hopkins University School of Medicine

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Rossella Verzulli
Sent: Monday, March 24, 2014 11:37 AM
To: statalist@hsphsun2.harvard.edu
Subject: st: pscore and attnd

Dear Statalist,

I am using pscore and attnd commands to find the nearest neighbour of each treated unit in my sample.

Could you please help me to find out how I can identify the group of matched control units? I would like to do some post-matching analysis after restricting my dataset to this sub-sample..

I will be very grateful for your advice.

Best,
Rossella


*
*   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/

*
*   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/