BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Correcting nonignorable nonresponse bias in turnout estimation usi
 ng callback data - Wang Miao (Peking University)
DTSTART:20260123T090000Z
DTEND:20260123T094500Z
UID:TALK241837@talks.cam.ac.uk
DESCRIPTION:Overestimation of turnout has long been an issue in election s
 urveys\, &nbsp\;with nonresponse bias or voter overrepresentation &nbsp\;i
 dentified as major sources of bias.However\, adjusting for nonignorable no
 nresponse bias is substantially challenging.Based on the ANES Non-Response
  Follow-Up study concerning the 2020 U.S. presidential election\, we inves
 tigate the role of callback data\, &nbsp\;i.e.\, &nbsp\;records of contact
  attempts in the survey course\, in adjusting for nonresponse bias &nbsp\;
 in the estimation of turnout.&nbsp\;We propose a stableness of resistance 
 assumption to account for nonignorable missingness in the outcome\,&nbsp\;
 which states that the impact of the missing outcome on the response propen
 sity is stable in the first two call attempts.\nUnder this &nbsp\;assumpti
 on and by integrating with covariates information from the census data\, &
 nbsp\;we establish &nbsp\;identifiability &nbsp\;and develop estimation &n
 bsp\; &nbsp\;methods &nbsp\;for &nbsp\;turnout.Our methods produce &nbsp\;
 estimates very close to the official turnout and successfully capture the 
 trend of declining willingness to vote as response reluctance &nbsp\;incre
 ases.&nbsp\;This work highlights the importance of adjusting for nonignora
 ble nonresponse bias and demonstrates the potential of widely available ca
 llback data for political surveys.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR