Urgent Multilevel Simple slope test s...
Message/Author
 Yi-Ying Chang posted on Wednesday, November 15, 2017 - 12:51 am
I'm much appreciated if you can advise me what the syntax for multilevel simple slope test in Mplus. Now I have two levels. bottom-up interaction. The interaction is high level moderator. The DV is high level variable. IV is lower level variable.
Thank you so much.
 Bengt O. Muthen posted on Wednesday, November 15, 2017 - 1:12 pm
See Cross-level interaction plots at our web page

http://www.statmodel.com/Mediation.shtml
 Maurits Masselink posted on Wednesday, July 25, 2018 - 3:45 am
Hi,

I have read different mplus forum posts and examples but unfortunately still do not understand how to do significant tests for the simple slopes in DSEM. I tried to work with model constraint and assigning labels to parameters, but this was not allowed. Could you point me to the right direction? To be clear, I do not only want to plot the interaction, but most importantly, I want to test te significance of the slopes.

In my model below, the SE_ACC effect is moderated by the between person variable BSE. What I would like to do is test slopes of high (+1SD) and low(-1SD) BSE.

DEFINE:
CENTER BSE (GRANDMEAN);

Model:
%within%
sSE | SE on SE&1;

SE_ACC| SE on Accept&1 ;
ACC_SE | Accept on SE&1;
ACC | Accept on Accept&1;

%between%
se accept SE_ACC on BSE ;
SE with Accept;

[BSE];

Maurits
 Bengt O. Muthen posted on Wednesday, July 25, 2018 - 4:06 pm
You could write on Between:

[SE_ACC] (a);
SE_ACC ON BSE (b);

Model Constraint:

Plot(sslope);
Loop (m,x,y,0.1);
sslope = a+b*m;

where x and y represent the lower and upper limits of your plot's x-axis and m represents values of the moderator BSE. For instance, x can be chosen as 1SD below the mean and y 1SD above the mean, where the mean refers to the sample mean of BSE and SD refers to the sample standard deviation of BSE.