BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Online Causal Inference Seminar: The Categorical Instrumental Vari
 able Model: Characterization\, Partial Identification\, and Statistical In
 ference - Richard Guo (University of Michigan)\, Yilin Song (University of
  Washington)
DTSTART:20260113T163000Z
DTEND:20260113T173000Z
UID:TALK240325@talks.cam.ac.uk
DESCRIPTION:We study categorical instrumental variable (IV) models with in
 strument\, treatment\, and outcome taking finitely many values. We derive 
 a simple closed-form characterization of the set of joint distributions of
  potential outcomes that are compatible with a given observed data distrib
 ution in terms of a set of inequalities. These inequalities unify several 
 different IV models defined by versions of the independence and exclusion 
 restriction assumptions and are shown to be non-redundant. Finally\, given
  a set of linear functionals of the joint counterfactual distribution\, su
 ch as pairwise average treatment effects\, we construct confidence interva
 ls with simultaneous finite-sample coverage\, using a tail bound on the Ku
 llback--Leibler divergence. We illustrate our method using data from the M
 inneapolis Domestic Violence Experiment.\nDiscussant:&nbsp\;Desire Kedagni
  (University of North Carolina - Chapel Hill) [Paper]\nFurther details abo
 ut the seminars are available on the OCIS webpage&nbsp\;
LOCATION:Discussion Room\, Newton Institute
END:VEVENT
END:VCALENDAR