Estimation I: Computational Companion

PLSC 30700 — University of Chicago

This site is a computational companion to the lecture slides for Estimation I (PLSC 30700), taught by Robert Gulotty at the University of Chicago.

Each chapter mirrors a lecture topic and demonstrates the key ideas in R. The goal is not to replace the slides or the textbook, but to give you a parallel way to build intuition: by computing, simulating, and visualizing the concepts.

1 How to use this site

Read alongside the lectures. Each chapter covers the same material as the corresponding lecture, but tells the story through code.
Run the code yourself. Copy-paste into R or RStudio. Modify the examples. Break things.
Focus on understanding, not memorization. The code is a tool for building intuition about what the math is doing.

2 Prerequisites

You will need R installed. The following packages are used throughout:

install.packages(c("ggplot2", "MASS", "car", "carData", "sensemakr",
                   "sandwich", "lmtest", "estimatr", "haven", "plm", "boot",
                   "AER", "gmm", "panelView", "fixest", "did", "fect"))

3 Chapters

Week	Chapter	Topic
1	1. Probability and Linear Algebra	Expectation, variance, the CEF, matrix algebra, and projection
1	2. The CEF and Best Linear Predictor	Why regression approximates the CEF
2	3. Multivariate OLS	Deriving and computing the OLS estimator
2	4. Sensitivity and Leverage	Frisch-Waugh-Lovell, partial R², influential observations
3	5. Efficiency and GLS	WLS, feasible GLS, method of moments
4	6. Small Sample Inference	Likelihood, the normal linear model, t- and F-tests
5	7. Probit and MLE	Binary outcomes, maximum likelihood estimation
5–6	8. Asymptotics	Large-sample theory, the delta method, and the bootstrap
6	9. Hypothesis Testing	F-test, test trinity, multiple testing, power
7	10. Instrumental Variables and 2SLS	Endogeneity, IV estimation, 2SLS, weak instruments, LATE
8	11. GMM	Moment conditions, efficient weighting, J-test, missing data
9	12. Panel Data	panelView, pooled/between/within, DiD, clustering, Arellano–Bond GMM
9	13. Fixed Effects and Modern DiD	RE, Hausman test, CRE, Callaway–Sant’Anna, fect, matrix completion

--- title: "Estimation I: Computational Companion" subtitle: "PLSC 30700 — University of Chicago" --- This site is a computational companion to the lecture slides for **Estimation I** (PLSC 30700), taught by Robert Gulotty at the University of Chicago. Each chapter mirrors a lecture topic and demonstrates the key ideas in R. The goal is not to replace the slides or the textbook, but to give you a parallel way to build intuition: by computing, simulating, and visualizing the concepts. ## How to use this site - **Read alongside the lectures.** Each chapter covers the same material as the corresponding lecture, but tells the story through code. - **Run the code yourself.** Copy-paste into R or RStudio. Modify the examples. Break things. - **Focus on understanding, not memorization.** The code is a tool for building intuition about what the math is doing. ## Prerequisites You will need R installed. The following packages are used throughout: ```r install.packages(c("ggplot2", "MASS", "car", "carData", "sensemakr", "sandwich", "lmtest", "estimatr", "haven", "plm", "boot", "AER", "gmm", "panelView", "fixest", "did", "fect")) ``` ## Chapters | Week | Chapter | Topic | |------|---------|-------| | 1 | [1. Probability and Linear Algebra](ch01-review.qmd) | Expectation, variance, the CEF, matrix algebra, and projection | | 1 | [2. The CEF and Best Linear Predictor](ch02-cef-blp.qmd) | Why regression approximates the CEF | | 2 | [3. Multivariate OLS](ch03-ols.qmd) | Deriving and computing the OLS estimator | | 2 | [4. Sensitivity and Leverage](ch04-sensitivity.qmd) | Frisch-Waugh-Lovell, partial R², influential observations | | 3 | [5. Efficiency and GLS](ch05-gls.qmd) | WLS, feasible GLS, method of moments | | 4 | [6. Small Sample Inference](ch06-small-sample.qmd) | Likelihood, the normal linear model, t- and F-tests | | 5 | [7. Probit and MLE](ch07-probit.qmd) | Binary outcomes, maximum likelihood estimation | | 5–6 | [8. Asymptotics](ch08-asymptotics.qmd) | Large-sample theory, the delta method, and the bootstrap | | 6 | [9. Hypothesis Testing](ch09-testing.qmd) | F-test, test trinity, multiple testing, power | | 7 | [10. Instrumental Variables and 2SLS](ch10-iv.qmd) | Endogeneity, IV estimation, 2SLS, weak instruments, LATE | | 8 | [11. GMM](ch11-gmm.qmd) | Moment conditions, efficient weighting, J-test, missing data | | 9 | [12. Panel Data](ch12-panel.qmd) | panelView, pooled/between/within, DiD, clustering, Arellano--Bond GMM | | 9 | [13. Fixed Effects and Modern DiD](ch13-fixed-effects.qmd) | RE, Hausman test, CRE, Callaway--Sant'Anna, fect, matrix completion |