Random coefficient prediction PreviousNext
Mplus Discussion > Dynamic Structural Equation Modeling >
Message/Author
 Megan Whelen posted on Thursday, June 13, 2019 - 1:33 pm
Hi,

I’m running a cross-classified DSEM model with continuous variables. I based my code on example 9.39 in the Mplus User Guide. In the example, the random intercept of y, random linear trend of y, and random slope of y on x are all regressed on the subject-level covariates. In my model, I only regress the random slope on my covariate, as that is the only effect I am interested in estimating. Are there any unanticipated consequences of leaving out the regression of the random intercept and random linear trend on the covariate?

Second, the unstandardized estimate I get when I regress the random slope on my covariate is .005. What is the interpretation of this number? I’ve noticed the options STD, STDY, and STDYX are not available when TYPE = CROSSCLASSIFIED RANDOM. Is there another way to get standardized estimates?
 Bengt O. Muthen posted on Thursday, June 13, 2019 - 5:07 pm
If you leave them out, you are saying that they are unrelated (uncorrelated) with the covariate. I would include them even if that's not of primary interest.

You can use a regular standardization approach because you know the variance of both the DV and the IV.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a public posting area. Enter your username and password if you have an account. Otherwise, enter your full name as your username and leave the password blank. Your e-mail address is optional.
Password:
E-mail:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: