Week-by-Week Outline

Below you will find a general outline of the topics we’ll cover in this course. All other materials regarding homework assignments, readings, and other information will be posted on the course website. Depending on how the class progresses, the order of these topics might change, so check regularly for an updated version of this outline:

Week Class Dates Topics
1 Aug 23 - Aug 25 Syllabus overview and course motivation; Multiple regression models: Quick review, statistical adjustment.
2 Aug 30 – Sep 1 Multiple regression models (cont.): Interactions in regressions; potential issues with regressions: outliers and collinearity.
3 Sep 6 – Sep 8 Regression models with discrete outcomes: logistic regression
Homework 1 due
Submit your group for the FINAL PROJECT
4 Sep 13 – Sep 15 Introduction to Causal Inference: Potential outcomes framework, the fundamental problem of causal inference, causal estimands, and study design.
5 Sep 20 – Sep 22 Randomized-Controlled trials: The gold standard? How can we use RCTs to answer causal questions? Limitations of RCTs. Internal and external validity. A/B testing.
Homework 2 due
6 Sep 27 – Sep 29 Introduction to Observational Studies: Can we use data outside experiments to make causal claims? Selection on observables. Matching and its comparison to RCTs. Overt and hidden biases. Propensity score matching. How does matching compare to regression adjustment?
Homework 3 due
7 Oct 4 – Oct 6 Observational studies: Using real-world variation to estimate causal effects. Natural experiments, Diff-in-Diff.
Submit your idea for the FINAL PROJECT (including data ideas)
8 Oct 11 – Oct 13 Observational studies (cont.): Regression discontinuity design (RDD).
Homework 4 due
9 Oct 18 – Oct 20 What about prediction? Overview of model checking.
Midterm exam due
10 Oct 25 – Oct 27 Building models: Bias-variance tradeoff; out-of-sample validation and cross-validation; Variable selection (e.g. stepwise selection).
11 Nov 1 – Nov 3 Model selection and regularization: Lasso and Ridge regression.
Homework 5 due
12 Nov 8 – Nov 10 Introduction to prediction models: K-nearest neighbors’ regression and classification.
Submit preliminary report for FINAL PROJECT
13 Nov 15 – Nov 17 Prediction models (cont.): Classification and regression trees (CART).
Homework 6 due
15 Nov 29 – Dec 1 Prediction models (cont.): Random forests, boosting, and ensemble methods.
Homework 7 due
16 December 5 Final Exam due
16 December 8 & 9 Final Project due & Final Project presentations

© Magdalena Bennett - licensed under Creative Commons.