Hi everybody,

I'm trying to estimate a random intercept and slope model, were I estimate the degrees of freedom with "dfmethod(satterthwaite)"

the code look like this:


mixed std_xenophobia_ij std_distrust_ij std_corruption_j age_ij women_ij eduyrs_ij std_contact_ij std_div_nbh_ij std_eco_depr_ij std_relig_ij eduyrs_pj muslim_j || cntry: , reml dfmethod(satterthwaite) dftable(pvalue) noheader noretable vsquish


But when I run it, I get the following error:


Performing EM optimization:

Performing gradient-based optimization:

Iteration 0: log restricted-likelihood = -37553.39
Iteration 1: log restricted-likelihood = -37553.39

Computing standard errors:

Computing degrees of freedom:
st_select(): 3200 conformability error
_mixed_ddf_get_X_matrix(): - function returned error
<istmt>: - function returned error
r(3200);



I have no problem using "dfmethod(satterthwaite)" in a more simple random intercept model that only include the main predictors (and no problem running the model without the dfmethod):


mixed std_xenophobia_ij std_distrust_ij std_corruption_j || cntry: , reml dfmethod(satterthwaite) dftable(pvalue) noheader noretable vsquish


Has anybody any idea of what is going wrong?


Best regards Anders