I have specified my model as a three-level multiple membership model (occasions, students, schools). The data are cross classified at level 3 (schools). Schools could be changed at 5 points in time.
However, the equation window does not show that my response variable (mathe) is modelled as a weighted sum of school effects. In fact, it does not differ from the notation for the simple baseline IGLS-model. What am I doing wrong? I attach the screen shot of the equation winow and the data structure (i = time, j = idaluni, k = s1, cons = ausw).
Many thanks in advance!
Claudia
wrong notaton for multiple membership models
-
ClaudiaSchuchart
- Posts: 5
- Joined: Fri Nov 08, 2013 6:20 pm
-
ChrisCharlton
- Posts: 1390
- Joined: Mon Oct 19, 2009 10:34 am
Re: wrong notaton for multiple membership models
I suspect that you need to change the notation to the single subscript version as described in chapter 16 of the MCMC manual. To do this click on the notation button and remove the tick for multiple subscripts.