Schedule
Week 1
- Thursday, January 23
Introductions, review of key ideas from STAT 155, why we need causal stories (graphs!)
Slides from class are available here.- After class:
Recommended reading: PRIMER, Chapter 1: Sections 1.1, 1.2
Required reading: PRIMER, Chapter 1: 1.4, 1.5.1
- After class:
Week 2
- Thursday, January 30
The mathematics (probability) of graphs- Before class:
Please watch the following two videos and/or read the corresponding sections of PRIMER. Answer the Moodle questions before class.
Video 1: Essentials of Probability, (slides)
Video 2: Key Structure in DAGs, (slides)
Reading: PRIMER, Chapter 1: Sections 1.3.1 to 1.3.5, Chapter 2: Sections 2.1 to 2.3
- Before class:
Week 3
- Tuesday, February 4
d-separation and its applications for understanding bias- Before class:
Reading: PRIMER, Chapter 2: Section 2.4. There is one associated Moodle question for this reading.
Please also read the article The Birth Weight “Paradox” Uncovered? available in the Readings section on Moodle. We will discuss in class on Tuesday.
- Before class:
- Tuesday, February 4
d-separation and its applications for understanding bias (continued)
No new reading. Work on Homework 1, due Thursday, February 13 in class.
Week 4
- Tuesday, February 11
The do-operator and estimating causal effects- Before class:
Required reading: PRIMER, Chapter 1: 1.3.6 and 1.3.7, Chapter 3: Sections 3.1 through 3.3 (skip 3.2.2 on Multiple Interventions). Answer the Moodle questions (under “The do-operator and estimating causal effects”) before class.
- Before class:
- Thursday, February 13
The do-operator and estimating causal effects (continued)
Week 5
- Tuesday, February 18
The do-operator and estimating causal effects (continued) - Thursday, February 20
Inverse probability weighting and structural models- Before class:
Required reading: WHATIF Sections 1.1 to 1.3 (Chapter 1: A definition of causal effect) and Chapter 12: IP weighting and marginal structural models.
No Moodle questions this week. Just read to familiarize yourselves with the ideas, which we’ll discuss in class this week. Chapter 12 is a little more technical - let the following questions guide your reading:- What is the concept of the “pseudo-population”? What is the nice characteristic that it has? How does IP weighting achieve that characteristic?
- How are stabilized IP weights different from ordinary (unstabilized) weights? Why would we want to use stabilized weights?
- What is a marginal structural model, and how are they fit? What parameters of these models are of interest?
- How is IP weighting used in dealing with censoring (loss-to-follow up / selection bias)?
- Before class:
Week 6
- Tuesday, February 25
IP weighting to account for selection bias - Thursday, February 27
No class - Capstone Days!
Week 7
- Tuesday, March 3
Causal discovery - Thursday, March 5
Causal discovery
Week 8
- Tuesday, March 3
Sensitivity analyses for unmeasured confounding - Thursday, March 5
Project work day
Spring Break: 3/16 - 3/27
Week 9: 3/30 - 4/3
If Google/YouTube access is a problem, all videos and slides are also available on Moodle in the “Videos & Slides” section. All Moodle questions are in the “Reading/Video Questions” section.
- Tuesday, March 31
Mediation analysis (video, slides)
Moodle questions: Mediation Analysis due Wednesday, April 1 at midnight (CST) - Thursday, April 2
Mediation analysis (continued)
Week 10: 4/6 - 4/10
- Tuesday, April 7
Randomized Controlled Trials (video, slides)
Quasi-Experimental Study Designs (video, slides)
Moodle questions: Study Designs due Wednesday, April 8 at midnight (CST) - Thursday, April 9
Randomized controlled trials and instrumental variable designs (continued)
Week 11: 4/13 - 4/17
- Tuesday, April 14
Interrupted time series and regression discontinuity designs - Thursday, April 16
Final exam work day
Week 12: 4/20 - 4/24
- Tuesday, April 21
Time-varying treatments - Thursday, April 23
Final exam work day
Week 13: 4/27 - 5/1
- Tuesday, April 28
Final exam work day - Thursday, April 30
Final exam work day