Hi all,

I have data with two levels: lets think of the classic example of students and schools.

When running the analysis in SPSS and selecting the Random Effects, we can select (we can or we have to?) the Subject combination

In Heck et al (2012) can be read: "Covariance structure. After specifying the random variables in the model, the user can select the Subject Combination (i.e., at what level the random effect will vary). In this case, the intercept is checked as a random effect that will vary across schools"

But the IBM Documentation says that SUBJECT is optional: https://www.ibm.com/support/knowledgecenter/SSLVMB_sub/statistics_reference_project_ddita/spss/advanced/syn_genlinmixed_random.html#syn_genlinmixed_random

So my question is: is it really optional? Or do we have to select the subject combination? What happens if we do not? I see the output changes significantly

All the best,

Alejandro

The reference I used is:

Ronald H Heck, Scott Thomas, Lynn Tabata - Multilevel Modeling of Categorical Outcomes Using IBM SPSS (2012).

Similar questions and discussions