R help - Jun 2007 - LanguageR pvals.fnc error message

Hi. I get an error message about not converging when I try and use the 
pvals.fnc from the languageR library. The LMER analysis worked fine (See 
below).

I am not an expert so I don't understand why the LMER worked but not the 
pvals.fnc

Any help gratefully received.

- Mike

   AIC   BIC logLik MLdeviance REMLdeviance
 -7324 -7254   3673      -7451        -7346
Random effects:
 Groups   Name        Variance   Std.Dev.
 Item     (Intercept) 5.3386e-05 0.0073066
 Subj     (Intercept) 1.8073e-03 0.0425128
 Residual             1.0314e-02 0.1015578
number of obs: 4363, groups: Item, 294; Subj, 38

Fixed effects:
               Estimate Std. Error t value
(Intercept)   5.848e+00  7.830e-03   746.9
PrevErr1      3.835e-03  3.808e-03     1.0
ExpOrd       -3.708e-05  9.107e-06    -4.1
pc4lex       -4.630e-04  1.681e-03    -0.3
pc4img       -7.519e-04  1.642e-03    -0.5
pc4nlen      -4.413e-03  1.608e-03    -2.7
pc4lsa        8.118e-04  1.648e-03     0.5
pc4ng       -5.183e-03  1.642e-03    -3.2
I(pc4nlen^2)  2.999e-03  1.451e-03     2.1

Mike Ford <mford <at> csl.psychol.cam.ac.uk> writes:
> 
> Hi. I get an error message about not converging when I try and use the 
> pvals.fnc from the languageR library. The LMER analysis worked fine (See 
> below).
> 
> I am not an expert so I don't understand why the LMER worked but not
the
> pvals.fnc
> 
> Any help gratefully received.
> 
> - Mike
  results of traceback() after the error?
  precise wording of the error message?
  any chance of sending to the list, or posting, a data
set that produces the error?  (see the posting guide ...)

  on the face of it is puzzling because pvals.fnc doesn't
seem to do any fitting, just post-processing of the lmer
object ...

  (by the way, that's "the languageR package" --
library(fortunes); fortune("Maechler.*package"))

  Ben Bolker