Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | natasha agarwal <agarwana2@googlemail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | st: Adjusting Standard Errors after Manually time demeaning the data. |
Date | Tue, 29 Jun 2010 09:11:05 +0100 |
Dear Everyone, I am estimating a fixed effect model on an unbalanced panel data. Since I wanted only robust standard errors and Stata 10 by default reports results where the Standard Errors are clustered at the panel identifier, I decided to manually time demean the data by using the "center" command. bysort number: center lnrval lnk lnw lnvfdi y14-y18, casewise prefix(c) reg clnrval clnk clnw clnvfdi cy14 cy15 cy16 cy17 cy18, vce(robust) Once I obtain the results, I wanted to adjust the Standard Errors for the degrees of freedom. I gave the following line of commands mat b=e(b) scalar vadj=e(df_r)/(e(N)-1-(e(df_m)) scalar vadj=e(df_r)/(e(N)-1-(e(df_m))) matrix V=vadj*e(V) eret post b V ereturn display I was not sure if this is correct since there seems to be no difference in the standard errors reported after ereturn display. Can anyone please help me with the same? Thanks Natasha * * 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/