Dear All,
I wondered if anyone could help with some latent class analysis (LCA) in R.
In particular, I wondered if someone could help regarding how to check the conditional independence assumption necessary for LCA, please?
My understanding is that there exists a log-odds ratio check (LORT) by Garret & Zeger (2000), but it doesn't seem to be a part of any R libraries I can identify.
https://www.john-uebersax.com/stat/condep.htm#detect
Any help on the topic would be greatly appreciated. I've attached some toy demo code from the poLCA package with K = 4.
Would appreciate all and any help, thanks.
install.packages('poLCA')
library(poLCA)
data(values)
str(values)
f <- cbind(A,B,C,D)~1
M2 <- poLCA(f,values,nclass=2)
M3 <- poLCA(f,values,nclass=3,maxiter=8000)
M4 <- poLCA(f,values,nclass=4,maxiter=8000)