Plots in GMM with categorical outcome PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Corinna posted on Wednesday, May 06, 2015 - 9:16 am
Dear Linda & Bengt,
I've got a four category outcome variable and want to plot the trajectories of this variable across four time points. I think I have to use the command "estimated probabilities". Here, I only get the trajectories for one category each. However, I would like to have like the "mean" trajectory of a class across all categories. If I use the command "sum of probabilities" I do not get a useful result. Can you tell me how I can plot the "mean probability" across all categories and time points? Is that somehow possible? Thank you very much
 Bengt O. Muthen posted on Wednesday, May 06, 2015 - 3:20 pm
I assume you are considering an ordinal outcome and not a nominal one. That is, you have it on the Categorical= list. For that case, I haven't seen the use of a mean probability and therefore we haven't included such an option.
 Corinna posted on Thursday, May 07, 2015 - 2:01 am
Yes, I have a categorical outcome (4 categories). I understand your point but is there any possibility to present the probability of a class across all categories and not for each category separately? (as it is in the command "estimated probabilities"?)E.G. through Excel or another program?
And another question: When I include the predictors in the model the option "estimated probabilites" for the plots is not available anymore. As I read in this forum, this might be due to the necessity of "numerical integration". Is there any way that I can still view the plots? Thank you very much for your help
 Bengt O. Muthen posted on Thursday, May 07, 2015 - 9:52 am
Q1. I don't recommend going that way. You will get critique from reviewers. Instead, consider a meaningful set of outcome categories and plot the sum of their probabilities, which is an option we offer.

Q2. Try the "Adjusted..." plot option.
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: