I want to estimate a CFA Modell with one factor. We used a booklet testdesign, so there are missings by design. The Factor is estimated with dichotomous items.
I used the PATTERN option but can't use the CATEGORIAL option.
When I run MPlus, this error message appears:
THE STANDARD ERRORS OF THE MODEL PARA- METER ESTIMATES MAY NOT BE TRUSTWORTHY FOR SOME PARAMETERS DUE TO A NON-POSITIVE DEFINITE FIRST-ORDER DERIVATIVE PRODUCT MATRIX. THIS MAY BE DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. THE CONDITION NUMBER IS 0.122D-18. PROBLEM INVOLVING PARAMETER 127.
THIS IS MOST LIKELY DUE TO VARIABLE LV1 BEING DICHOTOMOUS BUT DECLARED AS CONTINUOUS.
Not that I know. But it gets back to basic facts: For a dichotomous x the mean is p and the variance p(1-p) and therefore the two are mathematically tied. The first-order derivative matrix is therefore not full rank, resulting in the error message.