|
|
| Message/Author |
|
|
|
|
Hi
Can you please help me with specifying a simple model with 1 fixed intercept and 1 random intercept (corresponding to a cluster called pattern) as well as residual error. The outcome variable is loghdl. Is the following correct? The results give a correct estimate for the mean loghdl (i.e fixed intercept) but the random intercept for pattern and residual error don't match those of a conventional mixed model. Can you please suggest a better way of fitting this simple model?
CLUSTER IS pattern
ANALYSIS
TYPE IS TWOLEVEL RANDOM
MODEL: %WITHIN%
loghdl
%BETWEEN%
loghdl; ! I know this is not necessary
Thanks |
|
|
|
|
| You do not need RANDOM in this case to estimate a random mean. I suspect you are not using maximum likelihood in the other program and this is why the results do not match. If you have further questions, please send your Mplus input, data, output, the output from the other program, and your license number to support@statmodel.com. |
|
|
| Back to top |
|
|
