|
|
What does the H1 option do? Why are ... |
|
Message/Author |
|
Anonymous posted on Wednesday, August 17, 2005 - 9:56 am
|
|
|
1. Could you please explain what the H1 option does when type = missing? Comparing the output from when I do and don't specify H1, the only difference appears to be that the CFI, TLI and RMSEA are printed when H1 is specified. Is getting these stats the only reason to specify H1? Why aren't the CFI, TLI, and RMSEA printed when the H1 option isn't specified? 2. Could you please explain why intercepts are automatically computed under your FIML method by specifiying type = missing, but are not computed when type = general and listwise deletion is used? Thanks for the clarification! |
|
|
The H1 option estimates the unrestricited model. This is needed to compute chi-square and other fit measures. It is an option because it can be time-consuming to estimate. It's natural to include the means in the TYPE=MISSING; model. Results from a model with unrestricted means is the same as for a model without means. |
|
Back to top |
|
|