Causal Inference with Directed Graphs
A 2-Day Seminar Taught by Felix Elwert, Ph.D.
This seminar offers an applied introduction to directed acyclic graphs (DAGs) for causal inference. DAGs are a powerful new tool for understanding and resolving causal problems in empirical research. DAGs are useful for social and biomedical researchers, business and policy analysts who want to draw causal inferences from non-experimental data. The chief advantage of DAGs is that they are “algebra-free,” relying instead on intuitive yet rigorous graphical rules.
The two primary uses of DAGs are (1) determining the identifiability of causal effects from observed data, and (2) deriving the testable implications of a causal model. DAGs are also helpful for understanding the causal assumptions behind widely used estimation strategies, such as regression, matching, and instrumental variables analysis.
This seminar will focus on building transferable intuition and skills for applied causal inference. We start by introducing the essential elements for causal reasoning with DAGs and then use DAGs to discuss a range of important challenges in observational data analysis. Topics include: conditions for the identification of causal effects; d-separation; the difference between confounding, over-control, and selection bias; identification by adjustment; backdoor identification; what variables to control for in observational research; what variables not to control for in observational research; structural assumptions in regression and instrumental variables analysis; and recent work on causal mediation analysis.
Please note that this seminar will empower participants to recognize and understand problems and to spot fresh opportunities for causal inference. This seminar does not introduce new estimators and has no software component.
WHO SHOULD ATTEND?
If you want to understand under what circumstances you can draw causal inferences from non-experimental data, this course is for you. Participants should have a good working knowledge of multiple regression and basic concepts of probability. Some prior exposure to causal inference (counterfactuals, propensity scores, instrumental variables analysis) will be helpful but is not essential.
Since this seminar aims to strengthen your ability to think through causal problems we will work through numerous pencil-and-paper exercises. You will learn all necessary technical tools in this seminar. You do not need to know matrix algebra or calculus. There is no software component.
Schedule and materials
The class will meet from 9 am to 4 pm each day with a 1-hour lunch break from 12 pm to 1 pm.
The course will be taught in English.
Participants receive a bound manual containing detailed lecture notes (with equations and graphics), examples of computer printout, and many other useful features. This book frees participants from the distracting task of note taking.
Registration and Lodging
The fee of €915 covers all course materials.
Lodging Reservation Instructions
A block of rooms has been reserved at the Berlin Marriott Hotel, Inge-Beisheim-Platz 1, Berlin 10785 Germany, where the seminar is located, at a nightly rate of €155 for a room. It is advisable to check online for a lower rate, before booking with the group rate.
To make a reservation, you must call 0049 30 22 000 6300 or toll free 0049 800 18 54422 during business hours and identify yourself by part of the “Statistical Horizons” group. You can also book online by clicking here. The room block will expire when it is full or on February 19, 2016. Availability is limited to please book your reservation at your earliest convenience.
The city of Berlin levies a city tax of 5% on the room rate that is collected by the hotel. This tax does NOT apply to business related travel. To be exempt from the tax you must provide proof of business reason for the stay. This can be in several forms: Billing address is the customer’s company name and address, the invoice is settled through the customer company, or a set form.
1. Counterfactual causality
2. Directed Acyclic Graphs (DAGs)
b. Graphical display of the data generating model
c. The importance of causal assumptions
3. Associational implications of a causal model
a. Association vs. causation in DAGs
b. Three sources of association and independence (d-separation)
c. The difference between confounding and selection bias
d. Deriving testable implications
4. Graphical Identification Criteria
b. Control for confounding via adjustment
c. Adjustment and ignorability
d. Back-door identification
e. Front-door identification
f. Some helpful sufficient rules
5. Selection Bias
a. Post-outcome selection examples
b. Intermediate-variable selection examples
c. Pre-treatment selection bias examples
d. Why selection and confounding are distinct causal concepts
6. Graphical insights for common methods
a. Identification in matching and regression
b. Hidden causal assumptions in regression analysis
7. DAGs for mediation analysis
a. “Controlled” and “natural” effects
b. Identification requirements
“This course was an incredible eye-opener. I’m convinced that DAGs point to the future of sociological research using intuitive logics.”
Tarun Banerjee, Suny-Stony Book
“This course on causal inference is truly superb! Felix is an amazing teacher. I’ve sat in many sessions on DAGs and causal inference and this is the first time I can honestly say I understood much of the material. Felix’s ability to cross disciplines and help us understand the material is unique and extraordinary! Thank you!”
May Wang, UCLA
“This is a very good introduction to the use and interpretation of DAGs. I have learned a lot on how thinking about causality and all the pitfalls implied by looking for causal effects in empirical studies. DAGs are a great tool to improve understanding of these issues. The instructor explained the material very well.”
Jose A. Tapia, Drexel University
“The class focuses on issues that every researcher needs to understand. But consumers of research also need to know their concepts because, unfortunately, much of the literature (in medicine, at least) is composed of studies that ignore these principles. So it’s wonderful that Dr. Elwert is helping to disseminate this information. He is a great teacher.”
Richard Amdur, George Washington University School of Medicine
“This course should be required of all students and researchers who do any type of empirical research. The course in conjunction with Treatment Effect Analysis and Propensity Score Analysis are indispensable for any researchers wanting to make an impact in their applied research areas. Loved it!”
Lenny Lopez, Massachusetts General Hospital