Hello everyone!
I'm trying to construct several GLMMs with my thesis data and I can't solve this question. I have data on pollinator abundance and fruit set from plants sampled at six study sites and during two consecutive years. The problem is that not all individuals (plants) were sampled in both years.
My question is: How should i use "Year", as random effect or as fixed effect?
I constucted three candidate models and compared the AIC:
a) Response variable ~Predictor variables + (1 | year/plant) + (1 | site/plant)
b) Response variable ~Predictor variables + (1 | year) + (1 | site/plant)
c) Response variable ~Predictor variables * year + (1 | site/plant)
Option a) had the lowest AIC, but I read in some papers that is not correct use a variable with less than five levels as random effect...so, anyone could help me?
Thank you,
Raquel
Crawley (2002), Gelman (2005) A more useful definition of a random effect is a predictor variable where you are interested in making inferences about the distribution of values (i.e., the variance among the values of the response at different levels) rather than in testing the differences of values between particular
When deciding to model a factor as random or fixed, I make this question:
Are the differences among levels of the factor of biological interest and reproducible? In most cases "year" is a random factor. If you find differences between say 2000 and 2001 usually there is no clear biological reason that can explain the difference. Besides, unless one has a time machine, it is impossible to build the same model with different data from the same years. In contrast, "season" could be modelled as a fixed factor. There are biological reasons that could account for differences between say spring and autumn and one can test seasonal differences with different data.
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology