alex posted on Tuesday, January 20, 2004 - 8:02 am
I tried to compute the ML fiting function by computing the log likelihood of the covariance matrix assuming that it follows a Wishart's distribution but I cannot find the good result.???
bmuthen posted on Tuesday, January 20, 2004 - 9:04 am
The Mplus ML fitting function is given in (96) on p. 355 of the User's Guide. The Wishart-based likelihood differs from the Mplus likelihood (based on assuming normal outcomes) only by a factor corresponding to sample size n-1.