|
|
| Message/Author |
|
|
| Emma Twait posted on Monday, July 20, 2020 - 1:40 am
|
|
|
Greetings!
I am trying to perform a latent class analysis on binary and continuous variables, also with multiple imputation. I can understand the continuous means that are given per latent class; however, for the binary variables, logit regression is used. How can I also see the probability of 0/1 for the binary variables per latent class? Also, how can I get the latent class probabilities per data point in an imputation model? I read elsewhere that you can use the estimates from the pooled analysis, but not sure how this would work in syntax.
Thanks so much in advance! |
|
|
|
|
Q1: The probabilities for the binary variables are part of the output (scroll down below the regular estimates).
Q2: You have to run each imputed data set
separately. |
|
|
| Back to top |
|
|
