• DomeGuy@lemmy.world
    link
    fedilink
    English
    arrow-up
    23
    ·
    16 hours ago

    The first smell test for any survey is how would they possibly control for the non-response rate?

    Putting out a billboard to ask something like “what’s kind of makeup should a cracked egg try first” will get a bunch of recommendations and advertisment copy. But it wouldn’t tell you much about how many males wearing makeup are trans, enby, drag, or just wearing a costume. And noting at all about how many trans girls even try makeup at all.

    “Tell me your responses about how much HRT sucks” would, similarly, get you a dataset that’s highly distorted.