I have a quastion because I never found a model like this: Is it possible that two predictors on L2 (in my case personality traits of persons) can interact in predicting a random slope on L1 (states of negative affect on stress)?
I thought it should look like this:
USEVARIABLES = Stress NA Trait1 Trait2 INT;
WITHIN = Stress; BETWEEN = Trait1 Trait2; DEFINE: INT = Trait1*Trait2;
analysis: type = twolevel;
MODEL: %WITHIN% Slope| NA ON Stress;
%BETWEEN% Slope ON Trait1 Trait2 INT;
If it is possible, how can I model this and probe such an interaction? And is it necessary to center the predictors?