Trump nonresponse bias in polls

Homepage Forums Current Events Board Trump nonresponse bias in polls

Viewing 1 reply thread
  • Author
    Posts
    • #9026
      Mick1Mick1
      Participant

      In short, this article by Nate Cohn, the NYT chief political analyst, contends that Trump supporters either do not respond to political polls or if they do respond, they don’t exhibit their support for Trump. The article itself is from late 2022, right before the mid-term elections.

      In it, Cohn notes that in 2020, white registered Democrats were more than 20% likely to respond to surveys than white registered Republicans.

      Cohn also noted that the 20% disparity in 2020 grew to 28% by late 2022. Cohn’s conclusions was that there were not only too many Biden Democrats responding, but that there were likely too many Biden Independents and Biden Republicans responding.

      He has not updated his findings, but is considering doing so.

      Cohn is responsible for the methodology for the New York Times/Siena College polls.

      https://archive.ph/uUGWd

      Audaces fortuna iuvat

    • #9027
      Mick1Mick1
      Participant

      A pollster named David Miller points out that it requires an enormous number of outreaches to conduct a poll. In Florida, it took 90 dials to get one response. He refers to it as “convenience sampling of those consenting to give us their time and opinions.”

      In other words…polls are largely nonrepresentative, and therefore BS.

      How Non-Response Bias Can Affect Research Surveys

      Audaces fortuna iuvat

Viewing 1 reply thread
  • You must be logged in to reply to this topic.