Lagged effects PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 Martha  posted on Friday, March 21, 2014 - 7:26 am
I am trying to model data from a diary survey (5 days cluster in person). I would like to know how to study lagged effects.There is a variable (Y) predicted by previous (X) and same day variables (Z) and all variables are continuous. As long as I understand, it would be an equation similar to next one:
Y(ti+1) = B0j + B1J.X(ti)+ B2j.z(ti+1)+ Eij
Based on this, is it correct the next model?
i s | Y2@0 Y3@1 Y4@2 Y5@3;
y2 ON x1 z2;
y3 ON x2 z3;
y4 ON x3 z4;
y5 ON x4 z5;
Should I set y2 on x2[] y5 on x5 same-day effects to study lagged effects? Since I expect that Y changes across time are due to the IV, should I include regressions on the random i and s?
Thank you very much for your advice!
 Bengt O. Muthen posted on Sunday, March 23, 2014 - 12:15 pm
Your equation does not include the growth model i and s factors. But the Mplus input makes sense. Typically, time-varying covariates are not made to correlate with i and s (although one could argue that at least time 1 covariates could influence i and s).

Apart from this, the modeling choices are up to you; you may want to ask general modeling choice questions on SEMNET.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message