The Rogue’s Guide to Linear Models
Equations from District 7
Linear models are among the oldest and most fundamental tools in statistics, yet they are too often taught through examples of flowers and bees, genome sizes and plant heights, as if the mathematics existed in a vacuum, disconnected from consequence. This book rejects that premise. In an age where algorithms allocate healthcare, determine credit, and predict who is worth investing in, data analysis is no longer neutral, and pretending otherwise does students a disservice. Every regression line is drawn by someone, for some purpose, and increasingly those lines decide human outcomes. The Rogue’s Guide to Linear Models introduces simple regression, ANOVA, ANCOVA, and multiple regression with full mathematical rigor. Theorems are proved, assumptions stated, techniques demonstrated in reproducible R or Python code. But the methdology is embedded in StatPunk, a dystopian universe set in 2047 where data forensics specialists Beta and Bit use statistics to expose algorithmic injustice in the megacity of WaszKrak. Their investigations show what is truly at stake when a model treats some people with confidence and others with indifference. Because the next prediction an algorithm makes may decide whether someone receives treatment or is left behind, and that is precisely why understanding how these models work, and how they fail, has never mattered more. Without moralising, but with awareness.
