Chris Weber posted on Wednesday, July 29, 2009 - 7:12 pm
I have two outcome variables, which are both bimodal and continuous. I am trying to fit a two class mixture model:
NAMES ARE DOLKEPT EFFORT TOM d1 d2 d3 d4 d5 d6 d1xeyes d2xeyes d3xeyes d4xeyes d5xeyes d6xeyes;
USEVARIABLES ARE DOLKEPT EFFORT;
ANALYSIS: TYPE=MIXTURE; STARTS=100 20;
I get this error:
THE STANDARD ERRORS OF THE MODEL PARAMETER 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 -0.101D-16. PROBLEM INVOLVING PARAMETER 2.
I've tried specifying my own starting values,as well as increasing the number of random starts, but haven't had any luck. Am I missing something? Are there parameter restrictions that should be made?