BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal iCal API//EN
X-WR-CALNAME:Events items teaser
X-WR-TIMEZONE:America/Toronto
BEGIN:VTIMEZONE
TZID:America/Toronto
X-LIC-LOCATION:America/Toronto
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20190310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20191103T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:69f3202e4ca38
DTSTART;TZID=America/Toronto:20200113T100000
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20200113T100000
URL:https://uwaterloo.ca/statistics-and-actuarial-science/events/department
 -seminar-mamadou-yauck-mcgill-university
LOCATION:M3 - Mathematics 3 200 University Avenue West Room 3127 Waterloo O
 N N2L 3G1 Canada
SUMMARY:Department seminar by Mamadou Yauck\, McGill University
CLASS:PUBLIC
DESCRIPTION:SAMPLING 'HARD-TO-REACH' POPULATIONS: RECENT DEVELOPMENTS\n\nIn
  this talk\, I will present some recent methodological developments\nin ca
 pture-recapture methods and Respondent-Driven Sampling (RDS).\n\nIn captur
 e-recapture methods\, our work is concerned with the analysis\nof marketin
 g data on the activation of applications (apps) on mobile\ndevices.  Each
  application has a hashed identification number that is\nspecific to the d
 evice on which it has been installed.  This number\ncan be registered by 
 a platform at each activation of the application.\nActivations on the same
  device are linked together using the\nidentification number.  By focusin
 g on activations that took place at\na business location one can create a 
 capture-recapture data set about\ndevices\, or more specifically their use
 rs\, that \"visited\" the\nbusiness:  the units are owners of mobile devi
 ces and the capture\noccasions are time intervals such as days.  A new al
 gorithm for\nestimating the parameters of a robust design with a fairly la
 rge\nnumber of capture occasions and a simple parametric bootstrap varianc
 e\nestimator were proposed.\n\nRDS is a variant of link-tracing\, a sampli
 ng technique for surveying\nhard-to-reach communities that takes advantage
  of community members'\nsocial networks to reach potential participants. W
 hile the RDS\nsampling mechanism and associated methods of adjusting for t
 he\nsampling at the analysis stage are well-documented in the statistical\
 nsciences literature\, methodological focus has largely been restricted\nt
 o estimation of population means and proportions (e.g.~prevalence).\nAs a 
 network-based sampling method\, RDS is faced with the fundamental\nproblem
  of sampling from population networks where features such as\nhomophily an
 d differential activity (two measures of tendency for\nindividuals with si
 milar traits to share social links) are sensitive\nto the choice of a simu
 lation and sampling method. In this work\, _(i)_\nwe present strategies fo
 r simulating RDS samples with known network\nand sample characteristics\, 
 so as to provide a foundation from which\nto expand the study of RDS analy
 ses beyond the univariate framework\nand _(ii)_ embed RDS within a causal 
 inference framework and determine\nconditions under which average causal e
 ffects can be estimated. The\nproposed methodology will constitute a unify
 ing approach that deals\nwith simple estimands (means and proportions)\, w
 ith a natural\nextension to the study of associational and causal question
 s.
DTSTAMP:20260430T092606Z
END:VEVENT
END:VCALENDAR