Gregory posted on Tuesday, December 27, 2016 - 2:12 am
Hello, I have a question regarding how to specify a syntax for LPA with unequally spaced time points.
The model is aimed to identify different classes of trajectories that reflect the change in measurements over time. The time points are unequally spaced among subjects. Measurements are continuous variables. Data are wide.
I've seen the example (http://www.ats.ucla.edu/stat/mplus/faq/tscore.htm) of how to run a growth model with unequally spaced time points, but such approach does not accept 'tscores' for type = MIXTURE showing the following mistake: "TSCORES option is only available with TYPE = RANDOM."
Could you please help me with how to specify the syntax for my model?
TITLE: Entering data from .csv file DATA: FILE IS "E:/file.csv"; VARIABLE: NAMES ARE id v1-v40, t1-t40; USEVARIABLES ARE v1-v40, t1-t40; TSCORES = t1-t40; MISSING ARE ALL(999); CLASSES = CLASS(3); ANALYSIS: TYPE IS MIXTURE MISSING;
MODEL: %OVERALL% I S Q C | v1-v40 at t1-t40; %CLASS#1% Q@0; C@0; %CLASS#2% Q@0; C@0; %CLASS#3% Q@0; C@0;