www.cmm.bristol.ac.uk/forum

Hello,

is it possible to save the within-group residual variances per group (the sigma_sqrd_j's) in a twolevel intercepts-as-outcome-model? If so, how can this be done? I'd like to use the within-group variances for further analyses...

Many thanks for your response,
Nina

Hi Nina,
Not sure I understand your question with regard the per group statement. The model will have a variance at level 1 and level 2 and residuals at level 2 but there isn't a different variance for each group in this model.
Best wishes,
Bill.

Dear Bill,

sorry for my imprecise question - it is of course not the variance of the residuals per contextual unit, but the residual for contextual unit j.
For example, if there 50 contextual units, I assume that there will be 50 residuals, and it is these 50 residuals I would be interested in...

Many thanks for your reply!
Best,
Nina

If you use the Model->Residuals menu option then you can save the residuals for a chosen level into new columns. Is this what you are after?

...yes, thanks for this hint!
best
Nina