Course: Introduction to Causal Inference

Frequency:
Once per year, April through June/July

This course can be used as part of the research training for Epidemiologist (A/B) 

Causal Inference (CI) is a valuable method used to determine the causal effect of an exposure, such as treatment, on an outcome, like a disease. The goal is to understand and estimate the true causal effect of the exposure on the outcome while accounting for unobserved confounding factors, particularly in real-world settings, and using Big data in various fields including medicine, public health, and the social sciences. Yet, there is also a growing interest in gaining insights into causal relationships to drive evidence-based decision-making for policymakers.

Learning outcomes
1.  To understand the concepts of different methods to decide whether there is and what is the causal association of an or series of exposures with the outcomes and its impact on additive and a multiplicative scale.
2.  to learn methods used in causal inference like propensity score-based methods, inverse probability weighting,           instrumental variable methods, sensitivity analysis, mediation and moderation analyses, & time-dependent outcomes.
3.  To draw conclusions from the potential outcome's framework and causal diagrams under a variety of assumptions.

The format of the course (lectures, practical, self-study etc)
This 9-day CI course (one day per week) is designed to introduce a variety of causal inference tools. It follows a team-based learning (TBL) approach[1], with the following structure:
Phase I: advance assignment (out of class). Students receive a list of learning activities to prepare for the next phase. These activities may include readings, videos, tutorials, recorded lectures, etc.
Phase 2: readiness assurance (in class). Students complete a set of multiple-choice questions (MCQs) that focus on the concepts they need to master to be able to solve the Team Application (tAPP) assignment.
Phase 3: team application (starts in class): Students are presented with a scenario/vignette that is like the type of problem they may encounter in their careers. They are challenged to make interpretations, calculations, predictions, analyses, synthesize given information, and make specific choices from a range of options. They submit their work online and need to mark their peers using a peer marking system. The course runs over a period of ten weeks, thus 40 contact hours are dedicated to phase 2 and phase 3 of the TBL and 58 hours of self-study (assignments).

Course content
Week 1: Introduction to causal inference
Week 2: Causation and association
Week 3: Observations, modification and interaction
Week 4: Graphical representation theory I
Week 5: Graphical representation theory II
Week 6: Graphical representation – applications part I
Week 7: Graphical representation – applications part II
Week 8: Causal modeling
Week 9: The G-Formula and estimation
Week 10: Instrumental Design and Mendelian Randomization
Week 11: Assessment
Week 12: Resit

---

[1] A Parmelee, D. X., & Hudes, P. (2012). Team-based learning: a relevant strategy in health professionals’ education. Medical teacher, 34(5), 411-413.

Assumed pre-knowledge
To attend the course, students are expected to have completed a course in basic research design (or equivalent), basic statistics (or equivalent), and clinical epidemiology (or equivalent).

Reading materials
Hernán MA, Robins JM (2020). Causal Inference: What If. Boca Raton: Chapman & Hall/CRC and five to eight research articles.

Assessment
Student will complete an assignment about causal inference analyses in their own research of interest.
Minimal of 70% attendance is mandatory.

Contact
If you have questions on the course content, please send your email to Behrooz Z. Alizadeh.

EC (with exam)

3.5

Course coordinator

• Behrooz Alizadeh, PhD
  

Language

English

Target Groups

• PhD students
• ReMa students

Contact person

Back to listing