Back to Public SeminarsCausal Inference in R Using MatchIt and WeightIt: A Short Course
A 3-Day Livestream Seminar Taught by Noah Greifer, Ph.D.
This course offers an in-depth introduction to matching and weighting methods using R. Researchers use matching and weighting to estimate the causal effect of a treatment on an outcome — such as the effect of smoking on health or the effect of divorce on child outcomes — when assignment to the treatment is not random. Many of these techniques rely on traditional propensity scores, but the course will also cover newer techniques that do not. The R packages MatchIt and WeightIt (authored by the instructor) allow you to implement all of these techniques using a unified and easy-to-use syntax.
Matching and weighting are powerful, flexible methods that allow for the incorporation of substantive knowledge while providing transparency about the trade-offs that are often masked by other methods of estimating causal effects. Their outputs can be assessed and interpreted easily both by analysts and audiences, making them especially effective for medical and policy research.
Though the course is focused on these methods, other key scientific and methodological issues will be discussed, including communicating results, data visualization, and managing trade-offs between theoretical performance and interpretability.
Starting April 15, this seminar will be presented as a 3-day synchronous, livestream workshop via Zoom. Each day will feature two lecture sessions with hands-on exercises, separated by a 1-hour break. Live attendance is recommended for the best experience. But if you can’t join in real time, recordings will be available within 24 hours and can be accessed for four weeks after the seminar.
Closed captioning is available for all live and recorded sessions. Captions can be translated to a variety of languages including Spanish, Korean, and Italian. For more information, click here.
ECTS Equivalent Points: 1
More Details About the Course Content
This seminar is both conceptual and practical. We will briefly introduce the potential outcomes framework to motivate matching and weighting methods for causal inference. We will also discuss the conceptual differences between types of effects, including average treatment effects (ATEs) and average treatment effects on the treated (ATTs).
The course will then guide you through foundational methods like propensity score matching and inverse probability weighting to more modern methods that use machine learning and optimization (though prior knowledge of these topics is not required). We will discuss methods for evaluating the performance of matching and weighting methods, and end with estimating treatment effects, performing inference, and writing up the results of an analysis. You will get practical experience by working through exercises from the social and health sciences.
Computing
To participate in the exercises, we recommend you attend with recent versions of R and RStudio installed.
You need at least very basic experience with R. Users who are brand new to R can catch up by following this one-hour video on getting started with R before the course. Or, for more resources, you can check out our page on learning R.
Who Should Register?
This course is for anyone who wants to make better causal inferences from observational data. You should be familiar with linear regression. Some familiarity with logistic regression is also a plus.
Outline
Introduction to causal inference and matching
Theoretical background
-
- The potential outcomes framework
- Defining different treatment effects and estimands
- Assumptions for causal inference
Introduction to matching methods
-
- Exact and coarsened exact matching
- Mahalanobis distance matching
Propensity score matching and weighting
Propensity score matching
-
- The propensity score
- Propensity score subclassification
- Propensity score matching
- Customizing a matching specification: replacement, calipers, etc.
- Modern methods: genetic matching, full matching
- Implementation in MatchIt
Introduction to weighting methods
-
- Inverse probability weighting
- Targeting different estimands
- Machine learning methods for weighting: GBM
Covariate balance
Advanced weighting methods
-
- Optimization-based methods for weighting: CBPS, entropy balancing, energy balancing
- Implementation in WeightIt
Assessing covariate balance
-
- Assessing balance numerically and graphically
- Effective sample size
- Communicating balance
- Implementation in cobalt
Effect estimation and advanced topics
Estimating treatment effects
-
- Weighted difference in means
- Binary outcomes
- G-computation
- Accounting for uncertainty when matching and weighting
- Reporting the results of a matching or weighting analysis
Advanced topics (as time permits)
-
- Multi-category treatments
- Continuous treatments
- Missing data
Seminar Information
Wednesday, April 15 –
Friday, April 17, 2026
Daily Schedule: All sessions are held live via Zoom. All times are ET (New York time).
10:00am-12:30pm (convert to your local time)
1:30pm-3:30pm
Payment Information
The fee of $995 includes all course materials.
PayPal and all major credit cards are accepted.
Our Tax ID number is 26-4576270.