Testing Hypothesis in Mixed Linear Model

Hi All,

I am using a mixed linear model with robust standard errors. With robust SE, LR test is not valid. In such a case how can I test for the significance of the Fixed and Random effects? Wald test (or F-test in a finite sample) is an alternative but what about testing Variance Components? Is there any modified Wald Test (like the LR test with chi-square mixture distribution) as while testing for variance of RE, the test might be on the boundary of the parameter space.

Thanks,
sandip