• The Quuuuuill@slrpnk.net
    link
    fedilink
    English
    arrow-up
    61
    ·
    2 days ago

    in any poll you’re going to have like… 10-30% of people saying completely insane shit for a variety of reasons:

    1. not understanding the question
    2. not understanding the context of the question
    3. not understanding the context they live in
    4. willfully trying to fuck with the data set
    5. getting bored and no longer wanting to participate and just saying whatever gets the survey done faster

    generally speaking, anytime you get over 65% of people to agree on a topic, that is overwhelming support for that position on that topic

    • ArbitraryValue@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      21
      ·
      edit-2
      2 days ago

      Lizardman’s constant, an interesting short essay on this topic.

      Public Policy Polling’s recent poll on conspiracy theories mostly showed up on my Facebook feed as “Four percent of Americans believe lizardmen are running the Earth”.

      (of note, an additional 7% of Americans are “not sure” whether lizardmen are running the Earth or not.)


      The lesson from all three of the cases in this post seems clear. When we’re talking about very unpopular beliefs, polls can only give a weak signal. Any possible source of noise – jokesters, cognitive biases, or deliberate misbehavior – can easily overwhelm the signal. Therefore, polls that rely on detecting very weak signals should be taken with a grain of salt.