Dear Mandam or Sir,
I am doing some bootstrapping analusis of panel data model, especially
on bootstrapping the overidentifying test. While when I use the command as,
bootstrap, rep(99) cluster(ivar) id(newivar) group(ivar): xtivreg2 y x
(l.y=l2.y l3.y), fd kernel(truncated) bw(2) gmm robust
the J test statistic doesn't change at all as the asymptotic one. So I
doubt that command only bootstrap the standard erros rather than the
hansen test. In this case, I need to caculate the bootstrapped J
statistic by myself, so I need to get the access to all the 99
bootstrapped pseudo samples. How could I implement it in Stata?
And what's more, I looked for the block bootstrap in the Stata archive
while still haven't got a clear idea, could you please also help me
about it? Thanks!
best,
Grace
--
Miaomiao YAN
PhD. -st, College Assistant
Department of Economics
University of Vienna
Hohenstaufengass 9, A-1010 Vienna, Austria
Tel.: +43-1-4277 37458
--
Miaomiao YAN
PhD. -st, College Assistant
Department of Economics
University of Vienna
Hohenstaufengass 9, A-1010 Vienna, Austria
Tel.: +43-1-4277 37458
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
