st: Re: xtreg areg and R-square values

["Rodrigo A. Alfaro" <ralfaro76@hotmail.com>]

1. The R-squares between xtreg (fe) and areg are differt. I looked at the 
explanations on the stata faq. It seems to suggest that the xtreg command 
does not include the explanatory power of the dummy variable. However areg 
in my cases seems to provide a considerably higher R-square as compared to 
xtreg. my Areg r-squares are of the order of .25 and xtreg are .09. 
(overall) Does that mean that my indepent variables do not really contribute 
to explaining a lot of variation ?.  What is the correct thing to report.
Rod: areg uses dummy variables to compute R-2, xtreg shows 3 different 
R-2... there is not a unique measure, but usually within R-2 is reported.


I did see a query like this in stata listserv earlier. I dont think it 
provided a complete answer.
Rod: full description of the R-2 for xtreg is available in the reference 
manual. in the case of areg the R-2 is the same as reg.


2. the xtreg fe command in stata provides values of rho and values of 
standard deviation of the the random intercept. Why is that?. My basic 
knowledge is that fixed effect models dont have such values. For example 
areg does not provide these. What is the utility of these in the context of 
fixed effect panel data regression ?
Rod: in the reference manual (command xtreg) is explained this topic, after 
example 2.
*
*   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/