

Fisher Information Matrix in MCFA ana... 

Message/Author 


Hello, i am running a multilevel CFA, with 177 individuals nested in 34 teams. The output does not provide me with model test statistics, and the error message I get is the one below. I do not understand what to change in my syntax. Could you help me here please? Many thanks in advance, and best, Andreas THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO AN ILLCONDITIONED FISHER INFORMATION MATRIX. CHANGE YOUR MODEL AND/OR STARTING VALUES. THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NONPOSITIVE DEFINITE FISHER INFORMATION MATRIX. THIS MAY BE DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. THE CONDITION NUMBER IS 0.172D16. THE STANDARD ERRORS OF THE MODEL PARAMETER ESTIMATES COULD NOT BE COMPUTED. THIS IS OFTEN DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. CHANGE YOUR MODEL AND/OR STARTING VALUES. PROBLEM INVOLVING PARAMETER 36. 


Please send the full output and your license number to support@statmodel.com. 


Hello, I am running a MCFA, with 2,023 students nested within 93 teachers. I am running two different models using this sample, the first is a 3factor model and the other is a 4factor model, neither converges. The output does not provide me with model test statistics for either and I get the error messages below. I would sincerely appreciate any help. Do you have any advice or suggestions for me to help get these models to converge? Sandra THREE FACTOR MODEL ERROR MESSAGE: THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO AN ILLCONDITIONED FISHER INFORMATION MATRIX. CHANGE YOUR MODEL AND/OR STARTING VALUES. THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NONPOSITIVE DEFINITE FISHER INFORMATION MATRIX. THIS MAY BE DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. THE CONDITION NUMBER IS 0.273D15. THE STANDARD ERRORS OF THE MODEL PARAMETER ESTIMATES COULD NOT BE COMPUTED. THIS IS OFTEN DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. CHANGE YOUR MODEL AND/OR STARTING VALUES. PROBLEM INVOLVING PARAMETER 49. FOUR FACTOR ERROR MESSAGE: THE LOGLIKELIHOOD DECREASED IN THE LAST EM ITERATION. CHANGE YOUR MODEL AND/OR STARTING VALUES. THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO AN ERROR IN THE COMPUTATION. CHANGE YOUR MODEL AND/OR STARTING VALUES. 


We need to see the context. Please send those two outputs together with your license number to Support. 


Dear Sir or Madam, I have data from a diary study in which the observations on day level (n=747) are nested within persons (N=165). Now I´m trying to run multilevel CFAs to test for (a) the construct validity of my predictor variables on both levels and (b) the construct validity of my outcome variables measured on daily level. The tests for (a) all worked, but none of the tests for (b) – I always get the same error message like the users above me did: THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NONPOSITIVE DEFINITE FISHER INFORMATION MATRIX. THIS MAY BE DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. THE CONDITION NUMBER IS 0.355D19. THE STANDARD ERRORS OF THE MODEL PARAMETER ESTIMATES COULD NOT BE COMPUTED. THIS IS OFTEN DUE TO THE STARTING VALUES BUT MAY ALSO BE AN INDICATION OF MODEL NONIDENTIFICATION. CHANGE YOUR MODEL AND/OR STARTING VALUES. PROBLEM INVOLVING THE FOLLOWING PARAMETER: Parameter 143. But there seems to be nothing “special” about this parameter – and with other model specifications, just an other parameter is singled out as the possible cause of the failed estimation. I would be very glad for any tips on how to solve this problem! Thank you very much in advance! 


Please send your output to Support along with your license number. 

Back to top 

