Biology 708, McMaster University

Introduction

Generalized linear models generalize the idea of linear models (!) by adding:

• a family (the responses don’t have to be normally distributed)
• a link function (the linear predictor and the effect can be on different scales)

Like linear models, they are a very powerful set of tools depending on very specific assumptions, but at least they’re more flexible than linear models.

Class materials

• Main lecture notes
  • slide version
• More odds

Assignment

Make a generalized linear model for one or more of your hypotheses. Draw and discuss at least one of each of the following:

• diagnostic plot
• inferential plot (e.g., a coefficient plot, or something from emmeans or effects)

See the assignment instructions