Publications
Publications
- June 2021
- HBS Case Collection
Introduction to Linear Regression
By: Michael Parzen and Paul Hamilton
Abstract
This technical note introduces (from an applied point of view) the theory and application of simple and multiple linear regression. The motivation for the model is introduced, as well as how to interpret the summary output with regard to prediction and statistical inference. Using salary data from Glassdoor, the note provides a broad overview of correlation, simple linear regression, and multiple regression. Students will learn how to interpret regression coefficients and their corresponding p-values. The note also describes evaluation metrics such as r-squared and residual squared error. Finally, the note introduces students to diagnostic plots and reinforces the important concept that correlation is not causation. Throughout, the note demonstrates how these concepts can be implemented using the R statistical programming language.
Keywords
Linear Regression; Regression; Analysis; Forecasting and Prediction; Risk and Uncertainty; Theory; Compensation and Benefits; Mathematical Methods; Analytics and Data Science
Citation
Parzen, Michael, and Paul Hamilton. "Introduction to Linear Regression." Harvard Business School Technical Note 621-086, June 2021.