I tried to run the MIMIC model using MLR estimation. My factor model consisted of 4 factors(N=3107). I used integration=montecarlo(5000), but the model did not converge. The error message was: THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NON-ZERO DERIVATIVE OF THE OBSERVED-DATA LOGLIKELIHOOD. THE MCONVERGENCE CRITERION OF THE EM ALGORITHM IS NOT FULFILLED. CHECK YOUR STARTING VALUES OR INCREASE THE NUMBER OF MITERATIONS. ESTIMATES CANNOT BE TRUSTED. THE LOGLIKELIHOOD DERIVATIVE FOR PARAMETER 24 IS -0.20908945D+00. Copying the suggested starting values into my model command did not help the model to converge. Could you please advise on how I could get the model to converge? Is my syntax correct? Thanks very much.
Analysis: Estimator = MLR; INTEGRATION = montecarlo(5000); Model: Dep by bcesd03 bcesd06 bcesd09 bcesd10 bcesd14 bcesd17 bcesd18; Pos by bcesd04 bcesd08 bcesd12 bcesd16; Som by bcesd01 bcesd02 bcesd05 bcesd07 bcesd11 bcesd13 bcesd20; Int by bcesd15 bcesd19; Int @1;
Dep with Pos Som Int; Pos with Som Int; Som with Int; Dep Pos Som Int on sex bage bmwtdr bmtotal bcraven Bcode;
BCESD14 ON BAGE; BCESD17 ON SEX; BCESD08 ON BAGE; BCESD11 ON SEX; BCESD04 ON SEX;
Jon Heron posted on Friday, April 22, 2016 - 7:18 am
I think your factor model for INT needs additional constraints
We need to see your output to be able to judge. Send to Support along with your license number.
Daniel Lee posted on Friday, June 15, 2018 - 9:53 pm
Hi, I am trying to examine an interaction between a latent intercept term and an observed variable. When I run the model using Type=Random, I get the following error message when using xwith statement:
THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NON-ZERO DERIVATIVE OF THE OBSERVED-DATA LOGLIKELIHOOD.
THE MCONVERGENCE CRITERION OF THE EM ALGORITHM IS NOT FULFILLED.CHECK YOUR STARTING VALUES OR INCREASE THE NUMBER OF MITERATIONS. ESTIMATES CANNOT BE TRUSTED. THE LOGLIKELIHOOD DERIVATIVE FOR THE FOLLOWING PARAMETER IS -0.10091559D+03: Parameter 56, S1
I was wondering if you had recommendations for addressing these error messages. Thank you, as always.
The fact that the derivative is as large as -100.9 means that the maximization has failed. Perhaps it is hard to find a maximum of the logL for this parameter because the logL is flat - doesn't change much - for different values of the parameter, that is, the data set does not carry much information about its ML value. The maximization problem may also be for reasons discussed in the FAQ on our website:
TECH8 – negative ABS changes
Request TECH8 in the output to see if you have negative ABS changes. You can also see what the final value is for the "S1" parameter (parameter #56). Perhaps this parameter should be excluded from the model.
If this doesn't help, send your full output (with TECH1, TECH8) as an attachment to Support along with your license number.