[R-sig-ME] obtaining the range of a poly()'d fixed effect variable from an lmer object

Mike Lawrence Mike.Lawrence at dal.ca
Sun Dec 26 21:50:18 CET 2010

Hi folks,

I'm working on an update to the "ez" package, which has an
"ezPredict()" function to compute fixed effects predictions from an
lmer object. The version currently on CRAN breaks when the lmer object
was created with a formula in which poly() is used within the fixed
effects. I have a solution to this, but I need to know the range of
values for each fixed effect that is poly()'d, which doesn't seem to
be obtainable anywhere in the lmer object so far as I can tell (unlike
the case where a variable is numeric but not modified by poly(), in
which the range of the variable is easily obtained from the "frame"
attribute of the lmer object). Have I not looked hard enough?

For example, using the ANT data set from the ez package, I might
create the model:
fit = lmer(
    rt ~ (1|subnum) + poly(block,3)*poly(trial,3)*cue*flank*group
    , data = ANT[ANT$error==0,]

For ezPredict() to obtain predictions, I'd need to find the range of
the variables "block" and "trial". Thoughts?



Mike Lawrence
Graduate Student
Department of Psychology
Dalhousie University

Looking to arrange a meeting? Check my public calendar:

~ Certainty is folly... I think. ~

More information about the R-sig-mixed-models mailing list