On Thu, 29 Mar 2007, Talha Yalta wrote:
Thanks very much!!! Just in time before I need to do some
bootstrapping as a part of testing my results in my dissertation. It
works with the heteroscedasticity corrected OLS also right?
The "hsk" command? No, I'm afraid not. That's implemented using
an auxiliary regression followed by WLS, and supporting the
bootstrap for that case will require expanding the bootstrap code.
(The bootstrap code does not simply call the ordinary regression
functions over and over again: it is optimized for the repetitive
nature of the calculation.)
However, this could be done using a script, with the resample()
function.
Allin.