BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Epidemiological models for a Respondent Driven Sample - Märt Möl
 s\, University of Tartu\, Estonia
DTSTART:20080624T134500Z
DTEND:20080624T144500Z
UID:TALK12515@talks.cam.ac.uk
CONTACT:Nikolaos Demiris
DESCRIPTION:Researchers studying hidden populations (such as injecting dru
 g users) are often using chain-referral sampling schemes - snowball sampli
 ng or respondent driven sampling (RDS). Estimation and inference based on 
 a chain-referral sample is not trivial\, since the observations are likely
  to be correlated - friends tend to be more similar to each other than jus
 t two randomly selected persons from target population. There also exists 
 sampling bias\, because some people have higher probability of being sampl
 ed. Despite the obvious difficulties\, more than 100 studies in recent yea
 rs have used RDS. In the talk\, some likelihood-based methods will be prop
 osed for analysing datasets collected by RDS\, to make valid inferences fo
 r the parameters of interest.
LOCATION:Small Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR