Topics in Causal Inference

Beskrivning

In empirical settings without random assignment, researchers need to use alternative strategies to unveil causal relations. With a “natural” or quasi-experimental research design the researcher has control of the assignment to the treatment by using some criterion other than random assignment (e.g. eligibility thresholds, successive roll out of therapies, or praxis variation), and can use this to estimate the casual effect of the treatment. Such methods are commonly used in economic research.

The course will give an overview of quasi-experimental research methods for causal inferences, and will also introduce specific methods such as instrumental variables, regression discontinuity, difference-in-differences, and Mendelian randomization.
Josh Angrist (MIT) / Katrine Løken (University of Bergen) /Tove Fall (Uppsala University)

The workshop will offer all course participants an opportunity to present and discuss their current or recent research addressing causal questions in non-experimental settings, or to just participate in the discussions. The presentation and discussions should revolve around methodological aspects of the
paper.

Period	2017 juni 14 → 2017 juni 15
Typ av evenemang	Workshop
Plats	Uppsala, SverigeVisa på karta