BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Graphical models for causal reasoning in epidemiology - Vanessa Di
 delez\, University of Bristol
DTSTART:20080115T143000Z
DTEND:20080115T153000Z
UID:TALK8742@talks.cam.ac.uk
CONTACT:Nikolaos Demiris
DESCRIPTION:Graphical models are used to represent conditional independenc
 ies in multivariate systems\; they can facilitate model formulation\, reas
 oning and communication with subject matter experts as well as computation
 s. Here I will consider in particular their use for reasoning about causal
  inference in epidemiology\, where we typically want to investigate the ef
 fect of an intervention\, e.g. a public health intervention like banning s
 moking in pubs and restaurants\, on some health outcome. This task is chal
 lenging because most epidemiological studies are based on observational da
 ta and not on randomised studies. We therefore have to deal with the probl
 em that an observed association between exposure and outcome can be due to
  many other phenomena apart from an actual causal relation\, e.g. confound
 ing\, reverse causation\, selection
LOCATION:Large Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR