Estimated outcome probabilities PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 Gareth posted on Friday, February 10, 2012 - 3:20 am
For growth modelling with categorical observed variables using THETA parameterization, is it possible to calculate estimated outcome probabilities for chosen values of covariates on the slope factors? How is this achieved?
 Linda K. Muthen posted on Friday, February 10, 2012 - 10:02 am
In the regression of s on x, there are no probabilities involved because s is continuous. Is this the regression you are asking about?
 Gareth posted on Saturday, February 11, 2012 - 1:31 am
Not quite, I want to estimate the outcome probability for a chosen value of the covariate, for the mean slope value. So although s is continuous, I want to know the probability of this mean value when the covariate moves from 0 to 1.

The reason I want to do this is because probit coefficients are not very meaningful, due to the scale of the slope growth factor. My model can only be run in THETA mode.
 Linda K. Muthen posted on Saturday, February 11, 2012 - 4:18 pm
The regression of s on x is a linear regression. It is not a probit regression. The probit regression is u ON x. Is that what you want?
 Gareth posted on Sunday, February 12, 2012 - 3:50 am
I will e-mail my model to you, which should be easier for me to explain.
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: