Should I use fixed effects or random effects when I have fewer than five levels of a grouping factor in a mixed-effects model?
As linear mixed-effects models (LMMs) have become a widespread tool in ecology, the need to guide the use of such tools is increasingly important. One common guideline is that one needs at least five levels of the grouping variable associated with a random effect. Having so few levels makes the esti...
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PeerJ Inc.
2022-01-01
|
| Series: | PeerJ |
| Subjects: | |
| Online Access: | https://peerj.com/articles/12794.pdf |