print logo

MF9570 - New statistical methods for causal inference

Course content

New statistical models for causal inference are increasingly being used in epidemiology, clinical research and other fields. This course gives an introduction to some basic concepts and ideas in this area. Important concepts and methods that will be covered are: Directed acyclic graphs (DAGs), counterfactual causality, marginal structural models, direct and indirect effects, confounders, colliders, selection bias.

Learning outcome

New statistical models for causal inference are increasingly being used in epidemiology, clinical research and other fields. This course gives an introduction to some basic concepts and ideas in this area. Important concepts and methods that will be covered are: Directed acyclic graphs (DAGs), counterfactual causality, marginal structural models, direct and indirect effects, confounders, colliders, selection bias. These methods were developed over the last couple of decades, mainly by researchers at Harvard University (James M. Robins, Donald B. Rubin). The methods are part of a new approach to understanding how statistical analysis can form the basis of causal inference. In epidemiology and clinical research much knowledge about the causal effects of treatments and risk factors comes from statistical studies. The new tools give a much more precise way of approaching these issues. There is a great international interest in these approaches and we wish to make Norwegian reserachers acquainted with these developments.

Admission

Ph.D. candidates and students at the Medical Student Research Programme will get first priority to the course. The maximum number of participant is 35.

The StudentWeb is open for registration from 1st June to 15th August.

Registration for applicants without access to the StudentWeb from 1st June to 15th August.

Applicants will be notified by email 1 - 2 weeks after the final date for registration.

Prerequisites

Formal prerequisite knowledge

Introductory course in statistics

Recommended previous knowledge

Advantageous with some knowledge of logistic regression or Cox regression.

Teaching

The course will be taught 4th - 7th November 2013

The course is organized as full day teaching over 4 days, including lectures, exercises and discussions.

NB! You have to participate in at least 80 % of the teaching to be allowed to take the exam. Attendance at lectures will be registered.

Examination

A take-home exam will be given at the end of the course. Grading: Pass/fail.

Explanations and appeals

Students can request an explanation of their grades, and can also appeal against their grades or make a complaint about formal examination errors. Read more about explanations and appeals

Facts about this course

Credits

2

Level

PhD

Teaching

Every autumn

Examination

Every autumn

Teaching language

English

Semester pages

Teaching schedule, syllabus, examination date