[ADMB Users] quick, stupid question: retrieve standard errors for some parameters when bad hessian?

[dave fournier]

Just to round this out I modified the ADMB-RE fits from AGHQ to -is 100.
That is importance sampling with 100 points. This is useful because it 
can be used for
crossed random effects where AGHQ is impractical.
The type 1 error for ADMB-RE is 5.6%, near the theoretical value.
lmer is 32.4%.

ADMB_RE -is 100   0.056 250 0.994368 0.149674 0.143516
lmer    AGHQ=5    0.324 250 0.898267 0.160694 0.10748