Tech 14 Warning Message PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Berit Martin posted on Thursday, April 14, 2011 - 3:15 pm
I am attempting to run Tech 14 for an LCA with 8 categorical indicators (4 are 2-level, 4 are 7-level). I followed these guidelines suggested in the manual:

1. Run without TECH14 using the STARTS option of the ANALYSIS command to find a stable solution if the default starts are not sufficient.
2. Run with TECH14 using the OPTSEED option of the ANALYSIS command to specify the seed of the stable solution from Step 1.
3. Run with LRTSTARTS = 0 0 40 10; to check if the results are sensitive to the number of random starts for the k class model.

At step 3 I am getting the following Warning message:

WARNING: 1000 OUT OF 1000 BOOTSTRAP DRAWS DID NOT CONVERGE. THE P-VALUE MAY NOT BE TRUSTWORTHY. INCREASE THE NUMBER OF RANDOM STARTS USING THE LRTSTARTS OPTION.

I then successively increased the LRTSTARTS, and I am up to the point where I have LRTSTARTS = 0 0 1000 50, but I still receive the same error message. How far do I go before I give up? What does it even mean that the draws are not converging?

(Note: I also made LRTBOOTSTRAP = 1000)
 Linda K. Muthen posted on Friday, April 15, 2011 - 10:53 am
If number 1 has many replications of the bestloglikelihood and number 2 works, try number 3 with the default LRTSTARTS option. It may be that you need to use BIC and TECH11. TECH14 may have problems due to something about your model and data.
 Sung Kim posted on Friday, February 24, 2012 - 8:55 am
I am doing an LPA using 16 continuous indicators using 2,100 subjects. I'd like to get BLRT statistics (k =5) and keep getting warning despite large random starts. However, with smaller number of random starts the best likelihood was replicated in my LPA solutions. What is your recommendation? May I see the loglikelihood values of individual bootstrap draws?

ANALYSIS: TYPE = MIXTURE;
OPTSEED = 418686;
LRTSTARTS = 1000 20 1300 20;
! STARTS = 700 25;
! STITERATIONS = 20;
K-1STARTS = 1000 20;
! LRTBOOTSTRAP = 100;
PROCESSORS = 4;

WARNING: THE BEST LOGLIKELIHOOD VALUE WAS NOT REPLICATED IN 77 OUT OF 78
BOOTSTRAP DRAWS. THE P-VALUE MAY NOT BE TRUSTWORTHY DUE TO LOCAL MAXIMA. INCREASE THE NUMBER OF RANDOM STARTS USING THE LRTSTARTS OPTION.
 Linda K. Muthen posted on Monday, February 27, 2012 - 10:37 am
I would recommend:

LRTSTARTS = 0 0 500 200;

The 20 replications you have (for the K-class estimation in BLRT) is not enough.
 Sung Kim posted on Tuesday, February 28, 2012 - 1:33 pm
Thanks. So, I don't need to set the number of random starts and optimizations for the k-1 model in the LRTSTARTS option. The Mplus users guide says that one optimazation is carried out for the unperturbed set of starting values (the default for the k-1 class model). My question is how much we can trust the loglikelihood with unperturbed starting values? I specified like the following:

LRTSTARTS = 500 200 700 250;

Would that be too much for the k-1 class model?
 Linda K. Muthen posted on Wednesday, February 29, 2012 - 6:10 pm
Try

LRTSTARTS = 0 0 500 200;

and see how that works.
 Sung Kim posted on Thursday, March 01, 2012 - 9:14 am
Almost the same result (see below). What do you think?

WARNING: THE BEST LOGLIKELIHOOD VALUE WAS NOT REPLICATED IN 95 OUT OF 96
BOOTSTRAP DRAWS. THE P-VALUE MAY NOT BE TRUSTWORTHY DUE TO LOCAL
MAXIMA. INCREASE THE NUMBER OF RANDOM STARTS USING THE LRTSTARTS OPTION.
 Linda K. Muthen posted on Thursday, March 01, 2012 - 11:55 am
Please send the relevant files and your license number to support@statmodel.com.
 Chia Airu posted on Monday, April 10, 2017 - 9:11 pm
Hi Dr. Muthen

I am trying to run a latent profile analysis on 127 continuous variables using 900 subjects to fit 2 classes. However, i failed to obtain the BLRT statistics despite increasing the number of random starts using the LRTSTARTS option.

Could you please advise?

This is the input:

Variable:
names=x1-x127;
classes=c(2);
Analysis:
type=mixture;
estimator=mlf;
starts=0;
optseed=939021;
LRTSTARTS = 80 8 200 40;
Output: TECH14;

Error message:

THE LIKELIHOOD RATIO TEST COULD NOT BE COMPUTED. AN ERROR HAS OCCURRED DURING THE ESTIMATION OF THE H0 MODEL WITH ONE LESS CLASS.

INCREASE THE NUMBER OF RANDOM STARTS USING THE LRTSTARTS OPTION.
 Bengt O. Muthen posted on Wednesday, April 12, 2017 - 7:36 am
Please see our web note 14:

Asparouhov, T. & Muthén, B. (2012). Using Mplus TECH11 and TECH14 to test the number of latent classes. Paper can be downloaded from here. Mplus Web Notes: No. 14. May 22, 2012.
 Jie Gao posted on Monday, September 18, 2017 - 5:50 pm
I am running a LPA with 2 classes (4 continuous indicators).
ANALYSIS: TYPE= MIXTURE;
STARTS=2000 100;
STITERATIONS=100;
LRTBOOTSTRAP=500;
LRTSTARTS=1000 200 800 50;

MODEL: %OVERALL%
PRO5 PRE4 LOCO9 ASSE7;
%C#1%
PRO5 PRE4 LOCO9 ASSE7;
%C#2%
PRO5 PRE4 LOCO9 ASSE7;
OUTPUT: SAMPSTAT STANDARDIZED TECH11 TECH14;

But I got two following error messages:
WARNING: OF THE 494 BOOTSTRAP DRAWS, 256 DRAWS HAD BOTH A SMALLER LRT VALUE THAN THE OBSERVED LRT VALUE AND NOT A REPLICATED BEST LOGLIKELIHOOD VALUE FOR THE 2-CLASS MODEL.
THIS MEANS THAT THE P-VALUE MAY NOT BE TRUSTWORTHY DUE TO LOCAL MAXIMA.INCREASE THE NUMBER OF RANDOM STARTS USING THE LRTSTARTS OPTION.
WARNING: 6 OUT OF 500 BOOTSTRAP DRAWS DID NOT CONVERGE.
INCREASE THE NUMBER OF RANDOM STARTS USING THE LRTSTARTS OPTION.
I increased the number, but still got the similar warnings.Could you please advise what to do next? Many Thanks.
 Bengt O. Muthen posted on Tuesday, September 19, 2017 - 6:05 pm
See Mplus Web Note 14.

If this doesn't help, use BIC instead.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: