[R-sig-ME] questions on glmmADMB package

[Marcin Przystalski]

Dear All,

I've got several questions related to the package glmmADMB.



First. Correct me if I'm wrong. To find the estimates of unknown parameters
You apply the maximum likelihood method via Laplace approximation with the
restriction \alpha_{1}=0, i.e. that general mean can treated as the effect
of first object?

The second question is related with testing the hypotheses H_{0}:
\alpha_{k}=0. Do You use the following test statistic



z=\frac{\hat{\alpha}_{k}-\alpha_{k}}/\hat{\sigma}_{k}?



Further, You say that this test statistic under the null hypothesis has
N(0,1)?



I would be grateful for answering the questions.



Best regards


Marcin

	[[alternative HTML version deleted]]