\sigma^2 is the variance of the error term (or equivalently the variance of
the stochastic component in the alternate formulation) that we have to
estimate along with the \beta's We call it an ancillary parameter because
it does not vary over i and it is not of primary interest. We however do
have to estimate it in order to estimate our standard errors.
Brandon
On Mon, Jan 31, 2011 at 4:40 PM, Mabel Lana Botelho Abraham <mlba(a)mit.edu>wrote;wrote:
Hi,
I am working on the HW and I am unclear about 2.1. What do you mean by
sigma-squared ancillary parameter? Is this the variance or the intercept?
Also, should we be generating data for this probem?
Mabel
------------------------------
*From:* gov2001-l-bounces(a)lists.fas.harvard.edu [
gov2001-l-bounces(a)lists.fas.harvard.edu] On Behalf Of Molly Roberts [
molly.e.roberts(a)gmail.com]
*Sent:* Sunday, January 30, 2011 11:05 AM
*To:* Class List for Gov 2001/E-2001
*Subject:* [gov2001] lecture quiz
Hi everyone --
We noticed a few of you have not filled out the lecture quiz on the course
website. Although it is past the 8AM deadline, we still are interested in
understanding the points in the lecture you were confused about. If you
haven't filled out the lecture quiz, please do so as soon as possible!
Best,
Molly
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)lists.fas.harvard.edu
http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l