st: Literature reference for rho and LR-test in xtlogit, re

[s.vanbaal@arcor.de]

Dear Stata Users: 

I am looking for a "citeable" literature reference for a procedure explained in the PDF-manual. On page 227 of [XT] for Stata 11, the output from xtlogit is discussed. Here's a quote: 

"The output includes the additional panel-level variance component. This is parameterized as the log of the variance ln(s^2_u) (labeled lnsig2u in the output). The standard deviation s_u is also included in the output and labeled sigma_u together with r (labeled rho),

r = s^2_u / (s^2_u+s^2_e)

which is the proportion of the total variance contributed by the panel-level variance component. When rho is zero, the panel-level variance component is unimportant, and the panel estimator is no different from the pooled estimator. A likelihood-ratio test of this is included at the bottom of the output. This test formally compares the pooled estimator (logit) with the panel estimator."

I understand the procedure, but since I did not come up with it, I would like to cite someone. More specifically, I would like to cite a reference in my paper when introducing rho and explaining the LR-test result. 

Citing the Stata manual itself is possible, but far from optimal. I briefly checked the references given in the manual, but could not find anything fitting. 

Does anyone have a suggestion for a reference? I am open for suggestions from any discipline, although I have a certain preference for the econometrics literature. There may even be something in one of Wooldridge's books, but I couldn't find anything that fits the explanation quoted above. 

Thanks in advance! 
Sebastian van Baal
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/