raoul reulen wrote:
>>How do I perform a likelihood ratio test after having used Stata's mi
for imputation?
milrtest does not seem to work with Stata's mi, only with mim, but I
could be wrong.<<
As Jeph Herrin already noted, LR tests aren't really well defined for MI (as I recall), because the average of LR tests isn't the LR test. A Wald test is, in the sense that you can make a covariance matrix for parameters from MI.
(This is from memory, you should be sure to check references in the Stata manuals.)
*
* 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/
