I am trying to better understand the robust cluster variance
estimator, and the FAQ at
http://www.stata.com/support/faqs/stat/cluster.html
is helpful. But I have the following question:
For the unclustered heteroskedasticity-robust variance estimator, we
can take the corresponding formula in the FAQ, and write the standard
error of the slope estimator in a simple linear regression (i.e., one
regressor) as:
se(b_hat)= sqrt( Sum_i (x_i-x_bar)^2 * e_i_hat^2 ) / ( (Sum_i (x_i-x_bar)^2)^2 )
(Just as in the FAQ I am ignoring the degrees of freedom correction.)
Is there an analogous way to write the formula for the robust
*cluster* standard error of the slope term in a simple linear
regression?
Thanks,
David Ward
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/