in running a Generalized Linear Mixed Model through the wizard (i.e. not through syntax), the first dialog lets you specify a 'Subjects' variable. I always took for granted that such variable was then automatically fitted in the model as a random term, and even the SPSS manual seems to suggest so. See for instance the following bit of text:
However, I don't actually see any automatically created random blocks in the next dialog and if I manually add my subject variable (e.g. patientName) as a random term (in addition to specifying it as Subject in the first dialog), the output of the model is completely different.Help > Advanced Statistics Option > Generalized linear mixed models
Random Effects (generalized linear mixed models)
[omissis] ... By default, if you have selected more than one subject in the Data Structure tab, a Random Effect block will be created for each subject beyond the innermost subject. For example, if you selected School, Class, and Student as subjects on the Data Structure tab, the following random effect blocks are automatically created:
Thus, the question: should I manually add a random effect term for my subject variable or is it sufficient to specify it as Subject variable in the first dialog?
Thank you for helping!