Message/Author 


Hi  I am curious of the difference between the VuongLoMendellRubin LRT test versus the LoMendellRubin adjusted LRT test. What does the adjusted test adjust for, and is one preferred when making class enumeration decisions (obviously in conjunction with other criteria such as ABIC, BIC, and BLRT). Thanks! 


I think you should use the unadjusted version in general. The adjustment was made by them for the test to work better (better Type I error) for a certain situation. 


I have a twoclass GMM: %Overall% i s q  y1@0 y2@1 y3@2 y4@3; sq@0; c#1 on x; I was wondering what the k1 (oneclass) model would be in a Tech11 LMRLRT run? How can x be related to the categorical latent class variable in the oneclass model? Does this even make sense? 


X is not related – we remove the first class where all the beta coefficients are (not the last where all the coefficients are fixed to 0). To understand how tech 11 works you must carefully consider the reported loglikelihood value for the k1 class (that will be in this case a single class with no effect of X). That k1 class run should be run separately to ensure and verify that the test is for the correct k v.s. k1 classes. 


Thank you. So I simply leave out the x variable in the separate k1 run (this is a simple growth curve model). However, the LL from the separate run and the LL reported in TECH11 output for the 2class model do not match. Does x have to appear somewhere in the input syntax? This is what I did: 2 class model: Classes=c(2); Usevariables are y1y4 x; %Overall% i s q  y1@0 y2@1 y3@2 y4@3; sq@0; c#1 on x; (TECH k1 LL=26518.010) 1 class model: Classes=c(1); Usevariables are y1y4; %Overall% i s q  y1@0 y2@1 y3@2 y4@3; sq@0; (LL=26552.888) 


The code is correct and you should have obtained the same LL. You can use the data from User's guide example 8.1 and see that it gives the same LL. You can try using starts=100 with the 1 class model, or use some stricter convergence criterion such as mconv=0.0000001. If you still don't get the same LL send your runs and data to support@statmodel.com 

Back to top 