Launch Model with Categorical Contrasts PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 Seth Frndak posted on Friday, June 26, 2015 - 7:05 am
I am trying to determine differences in growth trajectory for three predefined groups. I have a working Launch Model and there is a significant group effect on the slope and frailty factors. I wanted to test differences between groups using one group as the comparison group. You can see that I have defined one group as the contrast group in the define statement. However, Mplus stops me from running this analysis due to missing values. However, this is exactly what I am testing and this is how a two-part model is set up. A large number of cases WILL be missing. What is the proper procedure here??

One or more variables in the data set have no non-missing values.
Check your data and format statement.
 Seth Frndak posted on Friday, June 26, 2015 - 7:06 am
if wpar==2 then earlyl=1;
if wpar==0 then earlyl=0;

if wpar==1 then midl=1;
if wpar==0 then midl=0;


!Launch Model;
!frailty factor for dichotomous variables;
frailty by L_D2@1 L_D3@1 L_D4@1 L_D5@1 L_D6@1 L_D7@1;

!Setting frailty intercept to 0;

!Setting threshold to 0 for wave 7;

ic sc | L_ln_C1@0 L_ln_C2* L_ln_C3* L_ln_C4* L_ln_C5* L_ln_C6* L_ln_C7@1;

!Constrain within-time intercepts to 0 for c variables for identification;
[L_ln_C1@0 L_ln_C2@0 L_ln_C3@0 L_ln_C4@0 L_ln_C5@0 L_ln_C6@0 L_ln_C7@0];

!Constraining variance of C1 to 0;

frailty with ic@0;
frailty with sc;
ic with sc@0;

frailty ic sc on earlyl midl ;
 Bengt O. Muthen posted on Sunday, June 28, 2015 - 10:32 am
Please send to support along with your license number and clarify if you refer to "two-part" growth modeling as described in the Mplus V7 UG.
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