Assumptions cross-classification models
Posted: Fri Apr 14, 2017 1:57 pm
Dear forum
I have a question regarding assumptions of cross-classification models. Do rules or assumptions exist about the minimum required crossing between classifications?
I’ll specify my question using an example: I’m using cross-classified models to investigate the effects of teachers, so I have a model of students within first grade and second grade teachers. There is no hierarchy between the first and second grade teachers, but there is also not so much crossing, since students mostly stayed in the same school and went on, with their classmates, to a second grade teacher. Comparing to other situations in which cross-classified models are used (f.e. students in neighbourhoods and schools), there is little crossing. Can this be a problem? Is a minimum amount of crossing required to have a good model? Or is it required that every student from a certain classification can be randomly assigned to the other classification? Or is this not at all a problem?
Many thanks in advance for your answer!
Griet
I have a question regarding assumptions of cross-classification models. Do rules or assumptions exist about the minimum required crossing between classifications?
I’ll specify my question using an example: I’m using cross-classified models to investigate the effects of teachers, so I have a model of students within first grade and second grade teachers. There is no hierarchy between the first and second grade teachers, but there is also not so much crossing, since students mostly stayed in the same school and went on, with their classmates, to a second grade teacher. Comparing to other situations in which cross-classified models are used (f.e. students in neighbourhoods and schools), there is little crossing. Can this be a problem? Is a minimum amount of crossing required to have a good model? Or is it required that every student from a certain classification can be randomly assigned to the other classification? Or is this not at all a problem?
Many thanks in advance for your answer!
Griet