Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

re: st: Propensity Score Matching

From   "Ariel Linden, DrPH" <>
To   <>
Subject   re: st: Propensity Score Matching
Date   Wed, 1 Feb 2012 10:21:19 -0500

If I understand your question correctly, you want to perform a test to
ensure that the covariates between the two groups "hedging" and "not
hedging" balance, ie., that they are comparable on observable baseline

If that is your question, then the easiest approach is to use -pstest- (a
user written program by the authors of psmatch2), right after running
-psmatch2- (also a user written program found on ssc).

If you are asking for something else, please clarify.

I hope this helps


Date: Tue, 31 Jan 2012 04:36:42 -0800 (PST)
From: vikramfinavker <>
Subject: st: Propensity Score Matching


I have a very basic question regarding propensity score matching. I am
trying to finding out the difference between firms that are hedging and
firms that are not on different factors. i want to know that the difference
in certain risk characteristics is just because of firms are hedging and not
because of any other factors. So, i want to identify firms which are
identical in all way (or nearest one) except for their choice of hedging.

my data looks like thisL

Year Hedging (1 = hegders, 0= non hedgers) risk1, risk2, sales, firm size,
r&d, leverage etc.

i have downloaded psmatch2, pscore. i did following things but i didn't get
what i am looking for:

. psmatch2 allheg10 vw1zscore vw1lnmva vw1mvgroslev vw1quickratio vw1asvol1,

Probit regression                                 Number of obs   =      
                                                  LR chi2(5)      =    
                                                  Prob > chi2     =    
Log likelihood = -1079.6428                       Pseudo R2       =    

    allheg10 |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
   vw1zscore |   .0077975   .0025215     3.09   0.002     .0028555   
    vw1lnmva |   .3208326   .0224338    14.30   0.000     .2768631    
vw1mvgroslev |    1.42135   .1899571     7.48   0.000     1.049041   
vw1quickra~o |  -.0653235   .0217963    -3.00   0.003    -.1080434  
- -.0226035
   vw1asvol1 |  -.1178394   .1869304    -0.63   0.528    -.4842162   
       _cons |    -3.1553    .315742    -9.99   0.000    -3.774143  
- -2.536457

. psmatch2 allheg10, n(1) pscore(vw1zscore)
There are observations with identical propensity score values.
The sort order of the data could affect your results.
Make sure that the sort order is random before calling psmatch2...

I want to see a comparison type of results for the nearest matched firms...


                     RIsk1           Risk2            
 Hedgers         0.000         0.000              
Non Hedgers    0.000         0.000             

So, i want to identify those firms which are identical in all way except for
their choice of hedging.


*   For searches and help try:

© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index