calculating effect sizes in ML-WIN


I have run multivariate multilevel regression using MLwiN 2.30. I would like to report effect sizes for the significant main effects as well as interaction effects. These involve both continuous as well as dummy variables.

Anybody know how to do this? And if it is not possible in Ml-Win, how to calculate effect sizes in another way?



Less is more. Stay pure. Stay poor.
Never used that program before, though I have a questions for you:

do you mean multivariate or multivariable? Former, multiple outcome variables; later, multiple predictor variables (independent variables).

I did indeed mean multivariate. By nesting multiple responses in participants you can analyze it in one model ánd this offers great flexibility than using Anova.