Pearson and LR chi-square test differed PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Yu-Chih Chen posted on Friday, April 01, 2016 - 12:29 pm
Hi: I am running a LCA model based on 9 ordinal variables. The model results showed that my chi-square values differ significantly in Pearson and LR chi-square. Given the df is the same, therefore, the significant test is varied (Pearson chi-square test is significant but LR test is insignificant). I am wondering which test I should trust and how can I fix this problem? Thank you so much!
 Linda K. Muthen posted on Friday, April 01, 2016 - 1:42 pm
When they do not agree, neither should be used.
 Yu-Chih Chen posted on Friday, April 01, 2016 - 1:58 pm
Thank you Dr. Muthen. Is that meaning my model is not trustworthy given these two tests disagree each other? How could I fix it, or is that I simply report other index (e.g., entropy, AIC, BIC)? Thanks!
 Linda K. Muthen posted on Friday, April 01, 2016 - 2:22 pm
It is more a shortcoming of the tests. Report another index.
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: