Multilevel logistic regression cannot compute an improvement -- discontinuous region encountered

[Juliette Corbi]

Hello,

I am trying to run a multilevel logistic regression with Stata. My outcome variable is a binary variable indicating the vote for radical right parties, level 1 is the individual unit, level 2 the country level. I am using the ESS dataset and I have decided to keep only 5 countries (keep if ctnry!=".." & ctnry!=".." etc)

In order to observe the Intracountry Correlation Coefficient, i have written this command

melogit votePRR || cntry :

estat icc

And I keep getting this message, after the calculation of 13 iterations

"cannot compute an improvement -- discontinuous region encountered" r(430);

Do you have an idea of what it means and what I could do?

Thanks a lot,
Juliette

[Carlo Lazzaro]

Juliette:
the message is telling that the mle cannot converge.
Tthe usual (partially correlated) recipes for this drawbacks:
1) add one predictor at a time and see when the problem starst to come alive;
2) run a more parsimonious model.

[Juliette Corbi]

Dear Carlo,

Thanks for your answer. I am not sure however that I understand what you mean with adding one predictor at a time and running a more parsimonious model, since I did not give any predictor in my command. Could you please tell me more about how I could proceed?

Thanks a lot,
Juliette

[Carlo Lazzaro]

Julitette:
I've tried (unsuccessfully) to replicate your problem with a Stata .dta:

. . use https://www.stata-press.com/data/r17/bangladesh
(Bangladesh Fertility Survey, 1989)

. melogit c_use|| district:

Fitting fixed-effects model:

Iteration 0:   log likelihood = -1297.1523 
Iteration 1:   log likelihood = -1295.4549 
Iteration 2:   log likelihood = -1295.4547 
Iteration 3:   log likelihood = -1295.4547 

Refining starting values:

Grid node 0:   log likelihood = -1277.4145

Fitting full model:

Iteration 0:   log likelihood = -1277.4145  (not concave)
Iteration 1:   log likelihood = -1270.5822 
Iteration 2:   log likelihood =  -1267.173 
Iteration 3:   log likelihood = -1267.0504 
Iteration 4:   log likelihood =   -1267.05 
Iteration 5:   log likelihood =   -1267.05 

Mixed-effects logistic regression               Number of obs     =      1,934
Group variable: district                        Number of groups  =         60

                                                Obs per group:
                                                              min =          2
                                                              avg =       32.2
                                                              max =        118

Integration method: mvaghermite                 Integration pts.  =          7

                                                Wald chi2(0)      =          .
Log likelihood = -1267.05                       Prob > chi2       =          .
------------------------------------------------------------------------------
       c_use | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
       _cons |  -.5382852   .0858361    -6.27   0.000    -.7065208   -.3700496
-------------+----------------------------------------------------------------
district     |
   var(_cons)|   .2495268   .0798179                      .1333023    .4670862
------------------------------------------------------------------------------
LR test vs. logistic model: chibar2(01) = 56.81       Prob >= chibar2 = 0.0000

. estat icc

Intraclass correlation

------------------------------------------------------------------------------
                       Level |        ICC   Std. err.     [95% conf. interval]
-----------------------------+------------------------------------------------
                    district |   .0704999   .0209614      .0389412    .1243258
------------------------------------------------------------------------------

.


.

Could you please share an example/excerpt of your data via -dataex-? Thanks.