MF9570 – Causal inference
Course description
Schedule, syllabus and examination date
Course content
Causal inference is the task of drawing conclusions from data about the effects of treatments and other type of interventions. In epidemiology and clinical research, as well as in many other fields, formal methods for causal inference play an increasingly central role. This course gives an introduction to basic concepts and ideas in this area.
Among the topics being covered are:
- randomization and target trials,
- counterfactuals and estimands,
- causal directed acyclic graphs (DAGs),
- methods for confounding adjustment,
- marginal structural models and time-dependent confounding,
- causal mediation analysis,
- causal inference in survival analysis.
The area of causal inference has over the last decades grown to be a very active area within statistics. Various new methods have been and are being developed, based on the seminal work by Donald Rubin, James Robins, Judea Pearl and others. This has led to new understandings of how statistical analysis is an integral part of causal inference and a continuously growing toolbox of methods for addressing causal questions.
In epidemiology and clinical research much knowledge about causal effects comes from statistical studies. The new tools give a more precise way of approaching these issues and can help researchers avoid common pitfalls. This course aim to make the participants acquainted with these methodological developments, both for the purpose of performing own research and for assessing the evidence from studies of treatment effects.
Learning outcome
- Understand the concepts of counterfactuals and causal estimands,
- Be able to use causal DAGs in practice,
- Be able to apply basic statistical methods for confounding adjustment,
- Understand the problem of time-dependent confounding and when more advanced methods are needed,
- Understand the challenges and possibilities of causal mediation analysis.
Admission to the course
Applicants admitted to a PhD programme at UiO apply to this course in StudentWeb.
Applicants who are not admitted to a PhD programme at UiO must apply for a right to study before they can apply to this course. See information here: How to apply for a right to study and admission to elective PhD courses in medicine and health sciences.
Applicants will receive a reply to the course application in?StudentWeb?at the latest one week after the application deadline.
Formal prerequisite knowledge
MF9130 – Innf?ring i statistikk / MF9130E – Introductory course in statistics or equivalent.
The course presupposes a thorough understanding of methodology as used in epidemiology and related fields.
Recommended previous knowledge
It is an advantage to have knowledge of logistic regression or Cox regression
Overlapping courses
- 4 credits overlap with MEDFL5570 – Causal inference.
- 2 credits overlap with MF9570 – Causal inference.
Teaching
The course is organized as full day teaching over four days, including lectures, exercises and discussions.
You have to participate in at least 80 % of the teaching to be allowed to take the exam. Attendance will be registered.
Examination
Home exam to be submitted four weeks after the course.
Language of examination
The examination text is given in English. You may submit your response in English, Norwegian, Swedish or Danish.
Grading scale
Grades are awarded on a pass/fail scale. Read more about the grading system.
More about examinations at UiO
- Use of sources and citations
- Special exam arrangements due to individual needs
- Withdrawal from an exam
- Illness at exams / postponed exams
- Explanation of grades and appeals
- Resitting an exam
- Cheating/attempted cheating
You will find further guides and resources at the web page on examinations at UiO.