Teaching

Graduate methods course readers for the University of Arizona, written to accompany the texts each course is built around.

How to use these. Each course has a reader — a book I wrote for the course — that you read alongside the textbooks listed with it, not instead of them. The readers restate the standard material in one notation, work every example in R, and show the code.

Take the two in order: POL 682 builds the linear model and the assumptions behind it; POL 683 picks up where those assumptions run out.

The readers are written in the methodology_ground_up repository, a shared teaching hub with Chad Westerland, and published here.

POL 682 · Linear Regression Analysis

Graduate methods · Fall · University of Arizona

First semester. The linear model from the ground up: least squares, the Gauss–Markov assumptions, inference, and what to do when the assumptions fail. Worked in R throughout, with a reproducible workflow.

Open the POL 682 reader

What the reader covers

  • Foundations: simple regression, BLUE, estimation, and the matrix form of the model.
  • Inference: variable selection, interactions, and interpretation.
  • When assumptions fail: heteroskedasticity, multicollinearity, autocorrelation, measurement error, and endogeneity.
  • Applications: panel data, plus a documented data appendix.

Companion texts — read alongside the reader

  • Fox, John (2016). Applied Regression Analysis and Generalized Linear Models (3rd ed.). Sage.
  • Fox, John, and Sanford Weisberg (2019). An R Companion to Applied Regression (3rd ed.). Sage.
  • Gelman, Andrew, and Jennifer Hill (2007). Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press.

POL 683 · Advanced Regression & Causal Inference

Graduate methods · Spring · University of Arizona

Second semester. What to do when the outcome is not continuous and the design is not experimental: limited dependent variables, simulation as a tool for understanding models, and the causal inference toolkit.

Open the POL 683 reader

What the reader covers

  • Inference and matrix algebra, revisited as the working language of the sequence.
  • Limited dependent variables: binary, ordinal, multicategory, and count outcomes.
  • Simulation for interpretation, prediction, and uncertainty.
  • Causal inference and model fit.

Companion texts — read alongside the reader

  • Long, J. Scott (1997). Regression Models for Categorical and Limited Dependent Variables. Sage.
  • McElreath, Richard (2020). Statistical Rethinking: A Bayesian Course with Examples in R and Stan (2nd ed.). CRC Press.
  • Gelman, Andrew, John Carlin, Hal Stern, David Dunson, Aki Vehtari, and Donald Rubin (2014). Bayesian Data Analysis (3rd ed.). CRC Press.