st: disattenuated regression

["'Alim Beveridge" <alimba@gmail.com>]

Hello,

I would like to run disattenuated regression (DR) in Stata. I have not 
found a package that will do this for me.
So I think I have to manually alter the variance-covariance matrix and 
run the regression on that altered matrix. I have never done anything 
like this before so any guidance would be really appreciated. I am 
running Stata 12 on Windows.

I found the following description of a procedure for manually altering 
the variance-covariance matrix (V): 
http://www.stata.com/support/faqs/statistics/instrumental-variables-regression/
In my case, I need to only correct the variances (for measurement 
error), the covariances will not change.
However, in my case the coefficients should change as a consequences of 
this correction, while in the example they do not. Does this mean I have 
to manually calculate the coefficients and update (b) too? Or is there a 
command I can use to estimate new coefficients based on the corrected V?
Can I use an approach similar to what is shown in the example? If not, 
what approach can I use?

Thanks for any help!
'Alim


*
*   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/