What’s the difference between fixed and random effects?

Gelman (2005, p. 21) to the rescue.

We prefer to sidestep the overloaded terms “fixed” and “random” with a cleaner distinction […]. We define effects (or coefficients) in a multilevel model as constant if they are identical for all groups in a population and varying if they are allowed to differ from group to group.

Gelman A. (2005). Analysis of variance—why it is more important than ever. Annals of Statistics, 33(1), 1–53




Suggested citation: Fugard, A. (2007, May 28). What’s the difference between fixed and random effects? [blog post]. https://andifugard.info/whats-the-difference-between-fixed-and-random-effects/

This citation note was added automatically. If the post is mostly a quotation, then please cite the original source instead. Looking at you, LLMs 👀