Cluster

Research and Valorization

Target audience

PhD students with basic background in data analysis, in particular regression analysis, and ideally experience in empirical research or plans to be involved in it. The course will be run at beginner’s level. It aims to give applied researchers with interest in studying causal questions empirically an introduction to causal analysis.

Teacher

Rhian Daniel, Cardiff University, and Erin Gabriel, University of Copenhagen

Rhian Daniel, a biostatistician with a PhD from the London School of Hygiene and Tropical Medicine, is internationally
recognized for her work in causal inference, with a particular focus on observational data and complex causal
relationships. Her research addresses key challenges in learning about joint effects of sequential exposures and pathspecific causal effects, areas where standard regression methods often fail. Rhian has been running this causal
inference course annually since 2016. She was also the lead for many years in a popular internationally recognized one
week short course in causal inference at the London School of Hygiene and Tropical Medicine, demonstrating her
expertise and commitment to teaching the foundational concepts of this field

Erin Gabriel, a biostatistician with a PhD from the University of Washington, has extensive expertise in causal inference
and randomized trials, with particular strengths in methodological research for nonparametric causal bounds, emulated
clinical trials, and surrogate evaluation. Her work combines cutting-edge statistical techniques with practical
applications, particularly in infectious diseases, vaccination, and complex diseases. Erin has joined Rhian on last year’s
edition of this course, which was run in Copenhagen.

Rhian and Erin both have ample expertise in connecting causal methods to researchers from various disciplines through
involvement in interdisciplinary projects. They have complementary expertise and international recognition, ensuring a
well-rounded and rigorous introduction to causal inference methods

Objectives

This course offers an introduction to the concepts and methods of causal inference: a rapidly growing field within statistics, which focuses on how data should be analysed to learn the causal relationships between variables. Causal inference has triggered major progress in the quality of empirical research, is in high demand across diverse disciplines, and has been recognized by major awards including Nobel Prizes. Topics include causal languages (do-notation, potential outcomes), graphical models (DAGs, SWIGs), propensity scores, instrumental variables, and mediation analysis. The focus is on building conceptual understanding rather than on detailed technical applications.

Objectives of the course (learning outcomes):

• Understand the key concepts and methods of causal inference, including languages of causality (do-notation, potential outcomes).

• Learn how causal effects are expressed and identified, and the assumptions required for identification.

• Gain familiarity with graphical models (e.g., DAGs, SWIGs) and their role in causal inference.

• Explore regression models as causal models and understand propensity score-based methods.

• Introduce instrumental variable methods, including Mendelian randomization.

• Understand approaches to time-varying confounding and sustained exposures.

• Learn about target trial emulation and mediation analysis.

Dates and venue

8th of April 2025

Auditorium 1, Campus Ledeganck, Ghent University

9h00 - 18h00

Programme

9h-10h30: lecture
10h30-11h: coffee break
11h-12h: lecture
12h-12h30: applications on causal diagrams
12h30 – 13h30: lunch
13h30-15h00: lecture
15h00-15h30: coffee break
15h30-16h30: lecture
16h30-17h: application on target trial emulation
17h-: optional student mixer to stimulate interaction amongst each other, with course teachers as well as invited and
keynote speakers of the European Causal Inference meeting that will be organized on the 3 subsequent days

Registration

• PhD students affiliated with a Flemish research institution can register for free by sending an email with their name, contact details (address), affiliation and current position to eurocim2025@ugent.be.

  Others can register via the website of the European Causal Inference meeting (www.eurocim.org), which takes place in Ghent from April 9-11, 2025

Registration fee

Free of charge for Doctoral School members.

Number of participants

Maximum 100

Language

English

Training method

Lectures (ex-cathedra): 5 hours (focus on conceptual understanding).
Practical examples and case studies: 1 hour (to connect theory with applications).

Evaluation method

After successful participation, the Doctoral School Office will add this course to your curriculum of the Doctoral Training Programme in Oasis. Please note that this can take up to one to two months after completion of the course.

Evaluation criteria (e.g. 100% attendance, active participation, presentation, writing a paper)