INTRODUCTION TO CAUSAL INFERENCE COURSE

Our introduction to causal inference course for health and social scientists offers a friendly and accessible training in the theory and practice of estimating causal effects in observational data. Delivered in-person over five-days by a team of experts, the course provides an unmissable introduction to this exciting and evolving aspect of contemporary data science.

COURSE CONTENT

  • The needs for, and principles of, quantitative causal inference
  • Counterfactuals, potential outcomes, and the Potential Outcomes Framework
  • Distinguishing (associative) prediction and causal inference
  • Introduction to causal directed acyclic graphs (DAGs)
  • How to draw and evaluate DAGs to aid with identifying and estimating causal effects in health and social science contexts
  • Covariate roles, the Table 2 Fallacy, and relevance for algorthmic interpretability
  • Classical confounding and propensity score approaches
  • Introduction to collider bias and selection bias, including Berkon’s paradox, the birthweight paradox, and outcome truncation bias
  • Reducing selection bias with inverse probability weighting
  • Exposure regimes and causal mediation analysis
  • Analysing multiple exposures and reducing time-varying confounding with G-methods
  • Including deterministic variables within DAGs to help with estimating and interpreting causal effects involving tautological associations, compositional data, and composite variables
  • Instrumental variable approaches, including Mendelian Randomisation
  • Threshold and discontinuity approaches, including difference-in-differences, interrupted time series analyses, and regression discontinuity designs

FURTHER DETAILS

FORMAT

Our introduction to causal inference course is a five-day in-person training event. It is not possible to attend remotely.

The course includes of a mix of face-to-face lectures, interactive activities, and question and discussion sessions. 

 

AUDIENCE

Our introduction to causal inference course is designed for applied health and social scientists.

Although we aim to make the material as accessible as possible, we recommend a working knowledge of linear regression modelling.

DATE & VENUE

Our next introduction to causal inference course will be held on 16-20th September 2024

The course will be held at Cloth Hall Court in the centre of Leeds, UK.

ATTENDANCE FEE

Our standard fee is £1095 + VAT (£1314), but a reduced fee of £795 + VAT (£954) is available for current students.

The fee includes attendance, access to all teaching materials,  and lunch and refreshments. Please note that the fee does not include accommodation.

OUR TEAM

Peter Tennant

PETER TENNANT

Peter is an Associate Professor of Health Data Science at the University of Leeds and a co-lead of the Alan Turing Institute’s Causal Inference Interest Group. He is a renowned public speaker and educator, and also internationally recognised for his research to adapt and introduce causal inference methods into applied health and social science research. He established the introduction to causal inference course in 2017.

Georgia Tomova

GEORGIA TOMOVA

Georgia is a Post-Doctoral Research Fellow at the University of Leeds and a co-lead of the Alan Turing Institute’s Causal Inference Interest Group. Her PhD research, adapting causal inference into nutrition epidemiology, attracted international praise and she is repeatedly invited to speak on the topic. She has been teaching on the introduction to causal inference course since 2019.

Laurie Berrie

Laurie Berrie

Laurie is a Post-Doctoral Research Assistant in Health and Wellbeing at the University of Edinburgh. Her PhD research, adapting causal inference into geographical research, led to seminal works on causal inference with compositional data and composite variables. She has been teaching on the introduction to causal inference course since 2017.

attend the course

Our September course is now fully booked, but some places may become available due to cancellations. If you are interested in attending, you can join our reserve list using the link below. Alternatively, you can also express interest in attending a future course (most likely in Summer 2025) via the link below.
.

TESTIMONIALS

Iain Brennan

IAIN BRENNAN

Professor of Criminology,

University of Hull

“Before attending the course, I had seen Peter (Tennant) deliver a short introduction to causal inference for the UK Reproducibility Network, and I had read ‘The Effect’ by Nick Huntington-Klein’. Attending the course really cemented my knowledge and allowed me to develop some more advanced thinking!

There was a real sense of openness and discovery in the group that was both forgiving of past mistakes and optimistic about new ways to do research with observational data. I really enjoyed the fact that those attending the course had a mix of disciplines, skills and knowledge and that the tutors were accommodating to all levels.

Six months later, I can honestly say that I have thought about this training and the things I learned every day since; it was truly revelatory and I believe it has allowed me to make a step-change in how I design research.

Within weeks of completing the training, I registered my PhD student on the course and I recommend it to anyone who will listen. Every quantitative researcher using observational data will eventually need to learn causal inference, and this course is the place to do it!”

Naomi Miall

NAOMI MIALL

Research Assistant in Social and Public Health Sciences,

University of Glasgow

“I cannot recommend the course more highly; it offers a really accessible and comprehensive training .

Before the course I thought I was pretty familiar with causal inference methods and was expecting a useful refresher, but it actually helped me to developed a much deeper understanding .

The opportunity to discuss the concepts with  such a broad range of attendees was one of the most valuable parts of the course, and I gained a great deal of confidence in assessing the quality of causal studies.

Peter (Tennant) and Georgia (Tomova) are endlessly energetic, patient, and have clearly spent many hours refining how to communicate these concepts and methods.

I would recommend the week on the back of the content and teaching alone, but it is also a really fun and social week that seems to attract very friendly and interesting attendees.”

Festes Seitenverhältnis

NICO OCHMANN

Research Associate in Economics,

University of Manchester

“The beauty of this course is it provides a very intuitive understanding of the concepts involved in causal inference without getting lost in mathematical statistics.

Being an economist, I learned causal inference with econometrics jargon. However, the interdisciplinary approach taken by this course is both productive and exciting because it applies similar concepts across various fields that use different terminology.

I would recommend this summer school to anyone who is willing to leave the math aside for a few days and learn about difficult concepts in a non-technical but challenging way.

The passion of the presenters for causal inference will be contagious. They love their stuff, so you will benefit a great deal”

Rob Aldridge

Rob Aldridge

Professor of Public Health Data Science,

University College London

“I had read quite a lot about causal inference before attending the course, but was struggling with understanding several topics (e.g. directed acyclic graphs) and how to apply them to my work.

I enjoyed so many things about the course, but if I had to choose one, it was the practical nature of the way causal inference was taught and the strong focus on how to apply this in your own research. The course was also extremely useful, the combination of extremely engaging lectures backed up with detailed and well written course notes that were far beyond anything I’ve got on a course before. I’ve been constantly referring back to them since taking the course and they’ve been invaluable.

Since attending the school, several members of our team have also been on the course, and they all enjoyed it as much as I did. We’re all learning together how to apply them and it’s been a great journey that started on the course.”

Lola Neufcourt

LOLA NEUFCOURT

Postdoctoral Researcher in Epidemiology & Public Health,

INSERM

“Before taking this course, I had some basic knowledge of causal inference, but this was mostly theoretical and the link between theory and practice was still missing.

I really enjoyed this course for several reasons: The course was very well prepared and structured, with many applications, real-life examples and activities to illustrate the concepts and make us think in practice.

There were also many opportunities to discuss with other participants from multiple disciplines and research areas about our own practices.

I really appreciated the involvement of the very dedicated and passionate instructors, who did not hesitate to spend a few extra hours answering (many) questions.

I would definitely recommend this course to anyone who wants to go into causality in depth and to rethink how to ask and answer research questions.”

Jessie Balwin

Jessie Baldwin

Postdoctoral Fellow in Psychology,

University College London

“This course is the best statistics course I have attended!

The tutors explain complex theories and methods in a clear and engaging way, making them very easy to follow.

Detailed notes and annotated code are provided, which are incredibly useful to refer to after the course has finished.

The tutors are also very supportive and show empathy for the challenges of learning about causal inference, which is rare for a course like this, and really helps with morale.

Finally, an added bonus of the course is that it’s also really entertaining – I’ve never laughed so much before when learning about stats!

In summary, this course is essential for anyone wanting to learn about causal inference.”

Sophie Pilleron

SOPHIE PILLERON

Postdoctoral Fellow,

University of Oxford

“I became interested in causal inference in observational data in early 2019 and took an online course on directed acyclic graphs, but I wanted to further improve my understanding and knowledge. I attended this course in September 2019 and the 5 days were truly eye-opening, once you attend this course you cannot read any epidemiological or biomedical papers in the same way.

Since the course, I use the knowledge that I learned every time I am thinking about a research question and every time I review a paper or grant proposal.

I am about to set up my own research group and will definitely be encouraging my new team members to attend this school. This course actively participates in improving the quality of health and social science research as a whole. It encourages researchers to think more about their research questions and how to answer them. I am still in the process of learning about causal inference, but I think that every researcher using observational data should take the type of training offered by this course.”

Olesya Ajnakina

Senior Research Fellow in Precision Medicine & Genetic Epidemiology,

King’s College London

“I had completed a PhD and three post-docs when I signed up to attend this course.  Without realising it, I had a very limited understanding of causal inference and the many problems that can arise from inaccurate data analyses.

The course was enlightening on so many levels. It clearly introduced the most important concepts and explained many biases and fallacies that are sadly not widely recognised or taught.

Since completing the course in September 2019,  I have been using all this knowledge and I have completely changed the way I carry out my analyses. I also recommend this course to all my students and colleagues; it is extremely important for high quality and accurate research.

Of particular note is Peter Tennant’s shining passion for the subject. I will recommend this course to anyone at any stage of their careers, who is keen to learn about how to do research accurately.”

Amiel Villanueva

Amiel Villanueva

MPH Candidate,

University of Glasgow

“I attended the course as a physician and Master of Public Health student familiar with DAGs and counterfactuals. The course not only deepened my existing knowledge, but also introduced several new concepts such as tautological associations and g-methods.

The content was excellent and we were given extensive lecture notes for reference. I also enjoyed the exercises in R , which demonstrated how to use the methods. But my very favourite part was having so  many opportunities to interact with the tutors during sessions, breaks and socials.

Peter (Tennant) and Georgia (Tomova)’s enthusiasm was palpable and they were regularly available to answer questions. I found the training directly useful for my master’s dissertation, but I believe it will be valuable to many working with quantitative data.

This course is for perfect for anyone serious about learning causal inference!”

Luisa Fassi

PhD Student,

University of Cambridge

“The Introduction to Causal Inference Course is truly a game changer for anyone looking to improve their methodological knowledge of causal inference.

Attending the school during the first year of my PhD has profoundly impacted my work. By applying a causal inference lens, I have now gained a much more critical view of both my research and papers in the field.

The content offered in the program is engaging and well-structured, covering in-depth essential topics like DAGs, collider bias, and the Table 2 fallacy.

I particularly enjoyed the blend of lectures, open discussions, and practicals, as well as the invaluable opportunity for personalised feedback through one-to-one discussions with the lecturers.

If you want to take your research to the next level, I can’t recommend this school highly enough!”

Alice Joules

AI Engineer,

IQVIA

“I work for IQVIA and my team focuses on building clinical prediction models for deployment into healthcare providers workflow.

The Introduction to Causal Inference Course provided key knowledge which helped shape our approach to developing fair and equitable clinical predictions models, understanding the utility and limitations of algorithm explain-ability techniques and conducting robust prospective validation studies of our algorithms once deployed into the healthcare system.

The teaching on this course is extremely clear, engaging and relevant and I highly recommend this course to all commercial organisations operating in the machine learning space. Thank you Peter!

These views are my own and not necessarily IQVIAs #IWork4IQVIA.”