www.cmm.bristol.ac.uk/forum

Hi all,
Scince it is not posible to do lr-test, how can i test (conclude that) that a logisitic two level-model is prefered to logistic with runmlwin?
Thanks for your help.

You could do a Wald test via the Stata -test- command (see http://www.stata.com/help.cgi?test). Examples of this can be found in the code replicating the logistic models chapter (9) of the MLwiN manual via -runmlwin- (http://www.bristol.ac.uk/cmm/software/r ... /examples/).