Mediation, Moderation & categorical I... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Tino Nsenene posted on Tuesday, October 18, 2016 - 4:29 am
Hello,

Mplus seems to be the only software able to estimate this model where a twolevel mediation (z-> x -> y) is combined with a cross-level interaction
(z-> slope). My example code is straightforward:

ANALYSIS:
TYPE = TWOLEVEL RANDOM;

MODEL:
%WITHIN%
SLOPE | y ON x;

%BETWEEN%
y ON x;
y ON z;
x ON z;
SLOPE ON z;

What is perhaps not so straightforward (& hopefully not outside the scope of this forum) is the question on the decomposition of the effects estimated with such a model:

Given that the slope of x varies between clusters and is predicted by z, how should one calculate the indirect effect from z on y?

Many thanks,
Tino
 Bengt O. Muthen posted on Tuesday, October 18, 2016 - 6:05 pm
If I look at the formulas correctly, I think the indirect effect from z to y via x is simply the usual product of slopes for y ON x and x on z. The random slope does not come into the picture for the indirect effect because there is not mediation on Within but on Between.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: