|
|
Three-level-Interaction in Multilevel... |
|
Message/Author |
|
|
Hi, 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? Kind regards Cosma |
|
|
Yes, this is possible. You have to also put Int on the Between list. You probe this between-level regression just like you do an ordinary regression with interaction. |
|
|
Thank you for ur answer! |
|
Back to top |
|
|