Predicting an outcome using LTA PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 Dylan Johnson posted on Wednesday, August 30, 2017 - 12:34 pm
I am currently looking at a two timepoint LTA with a 4 category latent varriable at each timepoint. Thus there would be 16 possible transition groups (4x4). I am looking to use these transition groups to predict my outcome of suicide attempts (dichotomous). Thus my model would look something like this:


x2 on x1;




My question pertains to setting up a reference group. I would like to calculate the odds ratios for each transition group relative to this transition group:


Is this possible in Mplus?
 Bengt O. Muthen posted on Wednesday, August 30, 2017 - 4:05 pm
Give parameter labels for each of the thresholds of [outcome$1]. Then use the Model Constraint command to express

- the probabilities of the outcome as function of the thresholds (see logit regression in UG chapter 14)

- the odds ratios as functions of the probabilities.
 Dylan John posted on Thursday, August 31, 2017 - 9:36 am
Thanks for getting back to me.

Are you saying that I want to add a parameter label to each of the 16 transition groups (a1-a16). Then in the logit regression statements I will specify the reference group as "a1"?
 Bengt O. Muthen posted on Thursday, August 31, 2017 - 4:59 pm
Yes, one label for each of the 16.

Then follow the style of e.g. slides

49-50 of short course Topic 6 on our website.
 Dylan Johnson posted on Friday, September 01, 2017 - 3:30 pm
Thank you!
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