|
Message/Author |
|
|
Is it possible to use Monte Carlo procedures in MPlus to examine the power of a hypothesis with 2 df? Suppose I have parameters i, j and k, and I am interested in the power to test the hypothesis that i=j=k. In observed data, one would set them equal and examine the decrement in fit with 2df. In the Monte Carlo module, you can look at the power of a 1 df test by defining new=i-k; in the constraint block and looking at the distribution of new across replications. Is there a way to do it for the two df case? Thanks, Eric |
|
|
I think Model Test gets Monte Carlo summaries - that gives a Wald test of constraints that you define. See the MODEL TEST command in the 4.1 UG. |
|
|
I don't see Monte Carlo summaries of Model Test statements.... should I send you the output? Eric |
|
|
Yes. |
|
|
Model Test produces the Wald Test fit that is shown in the Tests of Model Fit section at the top of the output. |
|
Back to top |
|
|