|
Message/Author |
|
|
Hi, I am running an LCA with 19 binary variables. When I run my LCA model the Peason Chi-square and Likeliehood ratio Chi square p-values equal 1. I have a few questions about this: 1.Is this because I have sparse data for some of the response patterns or is it something else? 2. Can I still use this model or does the pvalue of 1 make it invalid? 3. If I can still use the model are there any other statistics I use to assess model fit? Any help on this would be greatly appreciated. Thanks |
|
|
With over eight categorical latent class indicators the Pearson and Likelihood chi-squares are not reliable due to the very large multiway frequency tables that are being tested. The model may still be valid. You can use TECH10 to assess fit. |
|
Back to top |
|
|