Linear Modeling II and Logistic Regression

Lecture

• Slides
• In-class exercises
• Linear modeling supplement
  • An excellent overview of linear modeling, including model intepretation and fitting in R, written by Claus Wilke, PhD.

R content

• Broom vignette
• See Linear modeling supplement above as well

Linear models

• lm(Y ~ x + x1 + x2, data = data), with three independent predictors
• lm(Y ~ x*x1, data = data), with an interaction effect
• lm(Y ~ x + x1 + x:x1, data = data), with an interaction effect, written more verbosely
• lm(Y ~ ., data = data), to include all other columns as predictors

Logistic regression

• glm(Y ~ x + x1 + x2, data = data, family = 'binomial'), with three independent predictors
• glm(Y ~ ., data = data, family = 'binomial'), to include all other columns as predictors
• Code to extract relevant information for plotting:

model <- glm(Y ~ ., data = data, family = 'binomial')

plot.data <- tibble(x = model$linear.predictors,
                    y = model$fitted.values,
                    response = data$Y)