• where_am_i@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    18
    arrow-down
    1
    ·
    5 months ago

    This + confounding factors + a garbage can regression + a not so random small sample -> their findings are rubbish.

    • jacksilver@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      5 months ago

      Which is the same issue with many physcology and political survey-based studies. Take a survey that was horribly implemented, run some lightweight analysis (cause what can you really do with 20 respondents anyways), slap a funny name on it, and boom another piece of terrible research.