re: st: Synth (Abadia, Diamond, and Hainmueller) Placebo test graphic

["Ariel Linden, DrPH" <ariel.linden@gmail.com>]

Charlotte, 

When posting a question about a particular program, you need to indicate
where you got it from. In this case, -synth- can be currently downloaded
from http://www.mit.edu/~jhainm/Synth (Jens Hainmueller is now at MIT, but
was at Harvard when he originally wrote this program. So all recent updates
are found on his site).

As for your problem, it can be solved by adding in the -keep- option to the
end of the command (see below). This will generate a new output file after
each loop, with the annual values you are requestion. More specifically, you
need the following three fields: 
_Y_treated _Y_synthetic _time

After running the loop, I would merge the 4 files together (using the _time
variable as the anchor for the merge) so that you can plot the values for
each iteration. This is pretty straightforward.

I hope this helps

Ariel 

**** code *****
tempname resmat
        forvalues i = 1/4 {
        synth cigsale retprice cigsale(1988) cigsale(1980) cigsale(1975) ,
trunit(`i')  trperiod(1989) xperiod(1980(1)1988) keep(test`i')
        matrix `resmat' = nullmat(`resmat') \ e(RMSPE)
        local names `"`names' `"`i'"'"'
        }
        mat colnames `resmat' = "RMSPE"
        mat rownames `resmat' = `names'
        matlist `resmat' , row("Treated Unit") 
****end code****

Date: Wed, 11 Sep 2013 14:37:08 +0200
From: Charlotte Timmermans <timmermanscharlotte@gmail.com>
Subject: st: Synth (Abadia, Diamond, and Hainmueller) Placebo test graphic

Hello,

I am using the synthetic control method as developed by Abadie,
Diamond and Hainmueller with Stata 12.

I ran the method and got the results for my treated country (The
Congo) and for its synthetic counterpart made from a pool of 21
potential comparison countries.

I am now willing to implement the placebo test to verify the
significance of my results.

Following the synth help page provided by stata, I introduced the
following commands:

tempname resmat
       forvalues i =3D 1/21 {
       synth  rgdpch2  ini_gdp ci  LifeexpectancyBirth2
Populationgrowthannual av_pretreatment_gdp , trunit(`i')
trperiod(1965)
       matrix `resmat' =3D nullmat(`resmat') \ e(RMSPE)
       local names `"`names' `"`i'"'"'
       }
       mat colnames `resmat' =3D "RMSPE"
       mat rownames `resmat' =3D `names'
       matlist `resmat' , row("Treated Unit")

The program applied the SCM to every country in the donnor pool and
got the graph with the RMSPEs
I need to compute the graphic that computes the estimated effect
associated with every placebo run.
( A graph that shows the differences between per capita GDP paths (my
outcome variable) in every one of the 21 countries and their synthetic
counterpart. )
For example, in Abadie, Diamond and Hainmueller's paper about the
Effect of California=92sTobacco Control Program, shows the graph that
computes the placebo tests.

How should I proceed in Stata12 to compute this graphic? I am unable
to find the right comands.

Thank you for your time,

Best Regards.
*
*   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/