Dear Colleagues,
I would be appreciate any advice with regard to the handling of unequal
time interval in repeated measures regression with mixed models.
My objective is to examine effects of logging treatments (high, moderate,
low intensities) on tree species abundance over time (6, 10, 20, 36 months
after logging). I sampled plots in blocks imposed with the treatments .
Thus equal number of plots were randomly located in each of the blocks.
As an example, i modeled abundance in a repeated measures framework with
explanatory variables (treatment and time - of particular interest here)
with block/plot as the random effects
mymodel<-lme(abundance~treatment+treatment*month, random=~1 | block/plot,
data=sppghana, correlation=corAR(1))
Please, is this an appropriate usage of the AR(1) structure? Is there an
alternative way of handling the unequal time interval?
Akwasi