Welcome
Thank you for purchasing the MEAP edition of Regression: A friendly guide! Regression models are great options for analyzing relationships in data and for predicting outcomes, especially when interpretability is important -- explaining how to interpret and apply these models is therefore a basic goal of the book. After working through it, you’ll know what all of the lingo, notation, and output mean, you’ll know how to cut through the superficial hype surrounding regression and other data science tools, and you’ll know how to make sound, defensible data-driven decisions. Instead of just diving head-first into the theory, I’ll help you build up your intuition and knowledge incrementally, using carefully chosen datasets to motivate and illuminate things.
This book will teach you how to build, assess, and interpret regression models, from simple models with one predictor in Part 1 to complex models with many variables in Parts 3 and 4. You’ll learn to work with linear regression, logistic regression, regression for counts, robust regression and penalization, and generalized additive models. Along the way, you’ll also learn about simulation methods and modern tools like the bootstrap and cross-validation in Part 2.
The ideal preparation for this book is some previous experience with basic statistics: summary statistics (mean, median, standard deviation, quartiles), common distributions (normal, t, binomial), basic probability, common visualizations (histograms, bar charts), and significance tests. Don’t worry if these aren’t all familiar, but be prepared to do a little extra homework when they come up and let me know when you have questions. You should be comfortable programming in a data-friendly language like R, Python, or Julia; based on my own experience and preferences, I use R and tools from the tidyverse throughout the book, but the core statistical ideas are language agnostic..
As a MEAP reader, you can help me reach my intended audience -- you and others! -- by letting me know where I assume too much or too little about what you know, where my explanations could be improved, and where I’m being too long-winded or too terse. With that in mind, please don’t hesitate to share your comments, questions, and suggestions in the liveBook Discussion forum! I want this book to be a valuable resource on your journey to mastering regression, so your feedback will make a huge difference.
— Matthew Rudd