Hi Chiara,
There's no need to derive the distribution of the MLE -- what you
should do instead is calculate the Fisher information matrix and then
take inverse of that to get the variance. The standard error would
then just be the square root of that. I believe there's a discussion
of this in UMP (chapter 4, I believe).
hope that helps --
Maya
On Sat, Apr 3, 2010 at 2:49 PM, Chiara Superti <csuperti at fas.harvard.edu> wrote:
Hi Maya and Iain,
I have a quick clarification question about the problem set:
When you ask to find analytically the standard error of the MLE do you mean that you want
us to derive the distribution of the MLE?
If yes, since the derived MLE is a function of a r.v. that has a well known distribution
can we just derive the MLE distribution from that?
Thank you.
Chiara
_______________________________________________
gov2001-l mailing list
gov2001-l at
lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l