Session 1: Multiple linear regression review

Lecture

Learning objectives

1. identify systematic and random components of a multiple linear regression model
2. define terminology used in a multiple linear regression model
3. define and explain the use of dummy variables
4. interpret multiple linear regression coefficients for continuous and categorical variables
5. use model formulae to multiple linear models
6. define and interpret interactions between variables
7. interpret ANOVA tables

Outline

1. multiple regression terminology and notation
2. continuous & categorical predictors
3. interactions
4. ANOVA tables
5. Model formulae

Lab

Learning objectives

1. Load a tab-separated dataset into R
2. Create a simple scatterplot using ggplot2
3. Fit and analyze a multiple linear regression model
4. Compare two nested models using a nested Analysis of Variance (partial F test)

Exercises

1. Load the cholesterol.tsv dataset into R
2. Fit linear models with age, state, and interaction terms as predictors
3. Compare a simple linear regression to a multiple linear regression using Analysis of Variance partial F-test
4. Use backwards selection from a full model with interactions to choose the best prediction model