• bleistift2@sopuli.xyz
      link
      fedilink
      English
      arrow-up
      5
      arrow-down
      2
      ·
      2 days ago

      Was about to say that. It’s sad that your comment is the very last in this thread.

      • LillyPip@lemmy.ca
        link
        fedilink
        English
        arrow-up
        7
        arrow-down
        1
        ·
        edit-2
        2 days ago

        1: it’s not last, and 2: it’s not sad, because 3: people aren’t reading the source material. I love xkcd, too, but that doesn’t apply here.

        Just because results don’t match expectations doesn’t mean we should throw pies of satire in their face. That’s like the response in the OP of ‘no’. This is actually interesting.

        • Eheran@lemmy.world
          link
          fedilink
          English
          arrow-up
          5
          arrow-down
          1
          ·
          2 days ago

          The data does not support the conclusion. A simple “no” is okay. Take a look at these examples of regression. See how any one of the conclusions is absurd? Mind you the data in that example is far less random!

    • Don Piano@feddit.org
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      2
      ·
      2 days ago

      How do you think a case of “this explains some of the differences in the population, but not a lot” should look?

      And that looks potentially fine for an error bar. For a mean estimate, SE=SD/√N , so depending on what error band they used this looks quite plausible.

        • Don Piano@feddit.org
          link
          fedilink
          English
          arrow-up
          2
          ·
          1 day ago

          I recommend finding a different statistics teacher, preferably one who isn’t a comic and one who knows what the difference between a standard deviation, a standard error, and a 95% interval is. Those should not be too hard to find, it’s relatively basic stuff, but many people actually kinda struggle with the concepts (made harder by various factors, don’t get me started on the misuse of bar charts).

          • Eheran@lemmy.world
            link
            fedilink
            English
            arrow-up
            1
            ·
            21 hours ago

            I post the picture because it gets the point across, not because that is “my teacher”. The point is that you can choose smart any random regression function and they all fit just as “good”.