Page 1 of 1

Propensity Score Matching Algorithms

Posted: Thu Aug 06, 2009 1:05 pm
by gford
If you are looking for Propensity Score matching algorithms, see the paper at this link: http://www-wds.worldbank.org/servlet/WD ... ev0pdf.pdf. I have not implemented them yet, but plan to.

Re: Propensity Score Matching Algorithms

Posted: Tue Mar 06, 2012 6:08 am
by jleather43
Hi, I'm trying to implement this program but I have an error that says: 'missing values in @LOGL series at current coefficients at observations 2'. I know there are missing values in my data and i was wondering whether there was a way to get eviews to remove them as it does not seem to have done it automatically in this case?

Thanks

Re: Propensity Score Matching Algorithms

Posted: Tue Mar 06, 2012 10:28 am
by EViews Glenn
Set the sample so that the missings are excluded prior to estimation, as in

smpl if y<>na

If there are a number of series with NAs, put them in a group, say GRP1 then set the sample as

smpl if @rnas(grp1)=0

which sees whether there are missings in any of the series for each observation.

Re: Propensity Score Matching Algorithms

Posted: Wed Mar 07, 2012 12:42 am
by jleather43
Thanks