Multiple Predictors

Lecture 17

Dr. Elijah Meyer + Konnie Huang

Duke University
STA 199 - Fall 2022

October 26, 2022

Checklist

– Clone ae-16

– Review context of ae-15 for warm up question

Announcements

– Lab Feedback

– Project Repos Coming

Goals

– Finish categorical single predictor

– Model with multiple predictors

– Adjusted R-squared

Warm Up

Recall that we ended last class by modeling the relationship between a penguin’s body mass and the insland they were recorded on. Using the following output, write the estimated model below:

Further

– What if we want to account for both their flipper length and island?

– Would a different model explain more variation in the response body mass?

Modeling More Than one Variable

– Multiple linear regression is used to estimate the relationship between two or more explanatory variables and one response variable.

– we want to predict the value of a variable based on the value of two or more other variables

– “account for X1 and assess the relationship between X2 and Y” - main effects

– “does X1’s relationship with Y change based on X2”