

Weird df coming out from CFA tests 

Message/Author 

Tony Reed posted on Sunday, March 27, 2011  6:58 pm



Dear Drs. Muthen, I just finished running my model. And I am confused about the results of degree of freedom coming out of the test. Will appreciate if you can help me figure this out. I have a sample size of 612 (# of observations 610). In my measurement model, I have 43 dependent variables and 10 continuous latent variables. The results are as below: chisquare: 2265.31 df: 722 CFI: .918 TLI: .907 RMSEA: .051 SRMR: .048 All the item loadings are good. Nothing unexpected. My question is, why did I end up with a df of 722, which is even higher than the # of observations. Is it normal? I know there is a formula calculating the max # of df, which is t=1/2 *N*(N+1)(N= the number of dependent variables). In this case, t=946. So, right now the df=722 <t=946.> .9 would be acceptable though above .95 is ideal). Quite confused.... Also, is there any cut off value for X2/df ? Is it 3? Thanks!! 


Please send the output and your license number to support@statmodel.com. It is not possible to answer your question with the information that you give. 

Tony Reed posted on Tuesday, March 29, 2011  7:04 am



Will do!! Thanks, Dr. Muthen!! 

Back to top 

