Am 15.07.2015 um 13:39 schrieb JOSE FRANCISCO PERLES RIBES:
Following the first code-steps in Gretl provided by Professor Schreiber
and a SAS code provided by Professor Angrist (again, many thanks to both
Professors for your attention), I have performed a solution in R (sorry,
at the moment I'm more familiar with this language) to this issue.
No problem about using R. From what you sent me it looks as if Angrist
himself isn't really using this anymore -- I don't know why, but that's
not the best marketing he does for his own methods...
The code is weird, but is written with the aim to show the step-by-step
process and check the intermediate results.
If someone is interested in this topic, the data and my code could be
Not sure if this is urgent for you, but I would like to take a look at
this after the holiday season (realistically, after August). Feel free
to remind us if nothing happens until then.