Is level 2 variance significantly different from 0?
Posted: Sat Jan 18, 2020 1:57 pm
How do you work out if level 2 variance (the U-part) is significantly different from zero in the 2 level random intercepts only model?
That depends on which estimation method you are using but generally involves fitting the model with and without the random intercept and looking at some indicator of model fit - a LR test for ML methods or comparing DIC for MCMC. Suggest you look in the user guides for examples.
Bill.