|
|
LCA - Non-positive definite matrix |
|
Message/Author |
|
|
Dear Dr. Muthen, I used the following input statement for a latent class analysis: Missing are ALL (99); USEVARIABLES ARE S2WIT S2VIC S2NEI; CLASSES = c (5); ANALYSIS: TYPE = MIXTURE; LRTSTARTS = 0 0 500 200; STARTS = 2000 200; STITERATIONS = 40; OUTPUT: TECH1 TECH7 TECH11 TECH12 TECH14; Plot: type = Plot3; series = S2WIT (1) S2VIC (2) S2NEI (3); I received the following message in the output: 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 -0.481D-16. PROBLEM INVOLVING PARAMETER 2. I requested TECH1 in the OUTPUT, but I am not sure how to determine what is wrong with PARAMETER 2. Can you provide any suggestions? Thank you. |
|
|
Please send your output to Support along with your license number. |
|
Back to top |
|
|