|
Message/Author |
|
Jake Lant posted on Thursday, February 23, 2012 - 2:54 pm
|
|
|
I am trying to conduct moderation analyses (DV: continuous, IVs: Categorical, Continuous, Categorical*Continuous) on Mplus. The variables are non normal, and a good percentage of them are missing. This is for an assignment, so I am supposed to be getting these on the output: 3 estimates, S. E.s, and p-values (no constant). I need to rerun these and get the same numbers as the output. They are close on the ones I've tried but not quite there. I have tried to run different robust analyses just to get familiarized with the program but none of it seems to match the output I got (no ts p-values). Would anyone mind helping me figure out which commands would render an output with "estimate", "S. E.", "t", and a "p-value"? We were told we would need to use a robust estimator, which is why I've tried the robust analyses, by the way. Thank you in advance for any input that may clarify this. -Jake |
|
|
You should use DEFINE to create the interaction term. Your model would be y ON x1 x2 x1x2; All of the results are shown automatically in the output. Be sure you have the same estimator as the output and the same sample statistics. |
|
Back to top |
|
|