At 02:37 PM 4/23/2004 -0700, Maoyong Fan wrote:
use pdp_matching.dta, clear
...
Please find the data on:
http://are.berkeley.edu/~fan/imbens/
First off, the data set on this web page has a different name -
pdp_discuss.dta. So, is this the right file? Assuming it is, I ran your
do file in both Stata 7 and Stata 8, and no obvious differences stick out
at me -- the resulting file has 92 cases and 44 variables. But, perhaps I
overlook something. Could you either (a) provide the correct data file,
or (b) be more specific as to what it is that comes out
different? (Incidentally, I had to convert that file from Stata 8 to
Stata 7 format, so I am guessing it is not the right file, as you wouldn't
be able to run it in both versions of Stata.)
Also try running the program one line at a time and seeing when the
divergence begins. Then ask us why.
