Graphing growth cuves for different g... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Bill Dudley posted on Tuesday, September 23, 2008 - 2:52 pm
I am running quadratic latent curve models with time invariant predictors one is Binary (type of transplant allo vs auto). I want to plot the curves for transplant Allo = 0 vs Allo = 1.

I have
Analysis: type= random ;
MODEL: i s q|oe2@0 oe3@2 oe4@5 oe5@8 oe6@11 oe7@14 oe8@17 oe9@20;
i s q on allo;

The regression coefficients for slope and quad are statistically significant
I ON
ALLO 0.035
S ON
ALLO 0.102
Q ON
ALLO -0.004

Thus, the effect of allo =1 is to increase the slope by .102 and decrease the quad by .004.
Question
Are the regression coefficients (.102 and -.004) applied to the Intercept s that are provided as follows
I 0.062
S 0.105
Q -0.005
Thus, do the intercepts represent the intercept, slope , and quad for those with Allo = 0.
And are the curve parameters I would use for graphing the Allo= 1: Int = (.035+ .062) Slope = (.102 + .105) and quad = (-005 - .004) . OR do I look elsewhere and what TECH output do I need?
 Bengt O. Muthen posted on Tuesday, September 23, 2008 - 4:46 pm
You've got it right.
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: