topic Re: In Proc mixed, random effect-coefficient in Statistical Procedures

In Proc mixed, random effect-coefficient

dongsunk — Sun, 02 Jul 2023 13:03:50 GMT

I want a model that derives a random intercept from Proc mixed.

However, since the results are derived using SAS, coefficients are derived from Random Effects.

Is this the random intercept at level 2 I want?

Also, how to interpret this coefficient means subtracting the group mean from the overall mean?

Re: In Proc mixed, random effect-coefficient

jiltao — Mon, 03 Jul 2023 14:58:29 GMT

The Solution for Random Effects table shows the "deviation" for each subject from the overall mean. So you would need to add the fixed effect solution (intercept) to these deviations to get the subject-specific intercept estimate.

The following usage note provides some examples on this topic --

http://support.sas.com/kb/37109

Hope this helps,

Jill

Re: In Proc mixed, random effect-coefficient

dongsunk — Tue, 04 Jul 2023 11:30:11 GMT

Thanks for the very helpful answer.

For further questions in this regard,
Does adding the intercept for Solution for Fixed Effects to the coefficient for Solution for Random Effects give an estimate of the subject-specific intercept?
If not, do I need to add a constant term to the model coding of Random Effects?

Re: In Proc mixed, random effect-coefficient

jiltao — Wed, 05 Jul 2023 14:06:20 GMT

I am not sure I understand your question. Sorry.

Maybe you can provide your PROC MIXED program and relevant output? That might help me understand your question better.

Thanks,

Jill