It seems you have to estimate first your survival function and then compare
the values that this function take for the individuals in the treatment and
control group. This should be the relevant difference. The thing is how to
choose the individuals to be compared. You can use propesnity score
matching, mahalanobis distances or kernel matching (those are the matching
techniques use more often).
Take a look to psmatch and nnmatch commands.
Good luck
Manuel
----- Original Message -----
From: "Patricia Vera Rojas" <[email protected]>
To: <[email protected]>
Sent: Wednesday, August 17, 2005 8:26 PM
Subject: st: treatment effect
> Hi,
>
> I want to determine the impact of a labour market program
> in the employment results for participants(the treatment
> occurs at some time during the unemployment spell).
> Following Fredrikson and Johansson (2004)("Dynamic
> treatment assignment-the consequences for evaluation using
> observational data"), this effect can be estimated through
> the difference between survival functions for treated and
> not treated.
>
> After the logit/probit for getting the probability of
> participation in the program, what is the command for
> getting this difference?. Since "stslist" gives a table of
> probabilities of survival for the entire population and not
> for anyone, how can I get this difference?.
>
> Thanks,
> Patricia
>
>
>
>
>
> ______________________________________________
> Renovamos el Correo Yahoo!
> Nuevos servicios, m�s seguridad
> http://correo.yahoo.es
> *
> * 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/
