|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Mediation models w/ imputation |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
Marilyn Ahun posted on Thursday, November 15, 2018 - 8:32 am
|
|
|
I ran the following code to estimate an SEM model w/ 3 mediators using 10 imputed datasets. My output gives me the chi-square for all 10 datasets and the overall mean but the mean is 0 and so if my df - I don't understand where I went wrong. Script in input file: ANALYSIS: estimator = MLR; MODEL: pvic with inv atti; inv with atti; math on mds pvic inv atti ses iq verb stim func anti manx lpop fstat mage; pvic on mds ses iq verb stim func anti manx lpop fstat mage; inv on mds ses iq verb stim func anti manx lpop fstat mage; atti on mds ses iq verb stim func anti manx lpop fstat mage; model indirect: math ind mds; output: sampstat; stdyx; |
|
|
Maybe your model has not left-out arrows so that it is just-identified (saturated). |
|
|
Right. Thank you! |
|
Back to top |
|
|