Time Varying Covariate in ML Growth C... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Christian Beermann posted on Friday, October 01, 2010 - 6:17 am
Dear Mplus-Team and Forum,


In my study I want to control my measures at 6 timepoints in an Latent Growth Curve Model by a time-varying covariate.

Due to clustered data I will use either TYPE=COMPLEX or TYPE=TWOLEVEL. I tried both and I wonder how time varying covariates are handled in both cases.

TYPE = COMPLEX: Correlations with Intercept, Slope and the other Covariates are estimated (good fit, significant effects of the covariates)

TYPE = TWOLEVEL: Only the regression y1 on covariate1 is estimated (also good fit and significant effects of the covariates)

If I try to replicate the Twolevel Behaviour in the TYPE=COMPLEX Modell the fit is poor and the effects are not anymore significant.

It seems like there is anything I don't unterstand. Can somebody give me please a clue? Is that the right procedure to control for a time varying variable?

Thanks,
Christian
 Linda K. Muthen posted on Friday, October 01, 2010 - 8:17 am
We can't understand exactly what you are asking. Please send the three outputs referred to and your license number to support@statmodel.com.
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: