Due to Labor Day, we will not have in-person classes this week (all sections). For Monday sections, this recorded class will replace the final class on Dec 4th.
In this session, we will discuss potential issues with regressions, such as outliers. Finally, we will talk about models with discrete outcomes, and how we can tackle some of these issues.
Frost, J. (2019). “5 Ways to Find Outliers in Your Data” Note: You can skip the “Finding outliers with hypothesis tests” section
StatQuest. (2018). “Logistic Regression”. Note: Start at the 3:09 mark (Important: A logistic regression is NOT the same as a regression with a log(Y) outcome, as seen on the previous class)
Grace-Martin, K. (2018). “Why logistic regression for binary responses?"
Bellemare, M. (2013) “A Rant on Estimation with Binary Dependent Variables (Technical)"
Download the R Script for the “in-class” exercise here:
Here is the full R script for this class, with some additional questions:
You can also follow along the coding video to see how we answer some of the in-class exercises!Watch Video