Ordinary linear regression (OLR) assumes that response variables are continuous. Generalized Linear Models (GLMs) provide an extension to OLR since response variables can be discrete (e.g. binary or count). When both explanatory and response variables are categorical, it is more convenient to analyze data using contingency table analysis rather than using GLMs. Even though the two analyses are equivalent.
This course covers:
- What are GLMs? When should we use them?
- How GLM works.
- Categorical data analysis, including contingency table analysis, measures of association, tests of independence, tests of symmetry.
- How to use R to fit GLMs using real data and explain how we will interpret some of the output from the software.
There will be several examples to illustrate each type of models.