lme/lmer for drug effects on blood pressure

Thanks Andrew, much appreciated.

Andrew Dolman wrote:

Hi Matthew,

You haven't had any other replies so I'll just note that

lmer(pressure~drug*poly(timep,2)+(1|ID), data=drugdat)

looks OK to me.

If the random effects are not too complicated and you are interested in 
the effects of timep, you might like to compare results using the
SemiPar and mgcv packages -- both can fit GAMMs and the like (and make 
pretty graphics ;))

Cheers, David Duffy.