[R-sig-ME] LMER False Convergence (8)

[Daniel Anderson]

Dear all, 

I am trying to fit a cross-classified growth model with the lme4 package. The model is cross-classified because students moved between schools during the three years of the study (for a similar application, see Luo & Kwok, 2012). I have estimated the following model with both lme4 and the HLM software.

m.1<-lmer(MthScale ~ Clock + (Clock | mastid) + (Clock | School), data=dta2, REML = F, verbose = T)

where, 
MthScale is a mathematics achievement test, 
Clock is a time variable coded 0, 1, 2,
mastid = id for students,
School = id for schools.

When I fit the model I get the following error message "In mer_finalize(ans) : false convergence (8)". Yet, my results are quite similar to the results I obtained from the HLM software, which gave me no such error (see attached).

I have seen that some people have had success using David Hughes approach (http://davidhughjones.blogspot.com/2009/11/lme-false-convergence.html), but that did not work for me. I also read that it could be a result of the optimizer not reaching its predefined minimum, but that the result is likely to be a minimum (https://stat.ethz.ch/pipermail/r-sig-mixed-models/2011q1/015743.html). Finally, I tried fitting the following much simpler model, and still received the same error message:

m.2<-lmer(MthScale ~ Clock + (Clock | mastid), data=dta2, REML = F, verbose = T)

So, my questions are (a) are there any other ideas for getting the model to converge without the warning message, and/or (b) is it safe to continue with model building despite the warning?

Below is the output from the verbose = T argument. I've also attached a word document comparing the estimates between lme4 and the HLM software.

M.1<-lmer(MthScale ~ Clock + (Clock | mastid) + (Clock | School), data=dta2, 
+ REML = F, verbose = T)
 0:     754814.32:  1.01846 0.785304  0.00000 0.120908 0.0932292  0.00000
 1:     732113.86:  1.70314 0.384148 0.360769 0.600647 0.191005 0.0204139
 2:     723310.96:  1.68742 0.358177 0.164101 0.604350 0.200466 0.0158740
 3:     722264.35:  1.71511 0.320882 0.0856711 0.608439 0.211105 0.0130309
 4:     721675.24:  1.77901 0.280868 0.136016 0.614167 0.224845 0.0103563
 5:     721117.85:  1.90428 0.192638 0.0428979 0.608443 0.261661 -0.00547840
 6:     720620.64:  2.03047 0.133416 0.127071 0.663862 0.301313 -0.0567497
 7:     720235.93:  2.11910 0.209014 0.0676060 0.782523 0.253309 -0.0739699
 8:     720186.03:  2.08768 0.189996 0.0772091 0.804431 0.319605 -0.0664654
 9:     720183.13:  2.05956 0.147005 0.0884886 0.790482 0.261332 -0.0624898
10:     720173.52:  2.05849 0.146184 0.0791790 0.790283 0.263354 -0.0624119
11:     720168.15:  2.05605 0.139973 0.0832560 0.785274 0.265621 -0.0613035
12:     720166.44:  2.05367 0.134041 0.0812404 0.778806 0.267586 -0.0598906
13:     720166.44:  2.05367 0.134041 0.0812401 0.778806 0.267586 -0.0598905
14:     720166.44:  2.05367 0.134041 0.0812401 0.778806 0.267586 -0.0598905
Warning message:
In mer_finalize(ans) : false convergence (8)

Thanks,
--
Daniel Anderson
Research Assistant
Behavioral Research and Teaching
University of Oregon