Study
Longitudinal study of forest GrowthRates across time.
- 28 plots across 80 years, ~14 samples for each plot.
Current model:
lmer(GrowthRate ~ I(Year-1970) * factor(PlotType) + (1 + I(Year-1970)|Plot), data = dat)
Purpose
Determine the annual change in my response variable, GrowthRate.
- In other words, has GrowthRate increased through time?
Problem
I need to account for other variables by adding them to my model. Specifically, my plots vary in age, which likely plays a significant affect on the trend I see.
- Older plots are theoretically going to have different growth rates vs younger plots.
As a result, I need to account for the effects of age to focus in an actual temporal trend.
Question
Given that Year and Age are almost certainly strongly colinear, how could I incorporate Age into my model?
Thoughts: perhaps I could add an "initial age" variable to try to capture this effect? Would this work, or would this simply be repeating effects captured by my Plot ID structuring?
What's the best (or any appropriate) way to go about isolating the actual annual trend (i.e., the estimate for
Year) from trend effects due to plotage?
