• WhiskyTangoFoxtrot@lemmy.world
    link
    fedilink
    arrow-up
    3
    ·
    1 year ago

    As long as your sample represents an even cross-section of the population then it can be quite small, just about one order of magnitude larger than the result you’re reporting. For instance, if you’re reporting a percentage (i.e. a result out of 100) then you only need to poll about 1000 people, since cutting off one decimal place will be enough to account for any weird clustering in your sample. The total population being represented doesn’t matter; as long as your sample is truly chosen at random then all that matters is the sample size relative to the precision of the result.

    • FontMasterFlex@lemmy.world
      link
      fedilink
      arrow-up
      2
      arrow-down
      3
      ·
      1 year ago

      The sample size was 1229 people for this “survey”. With something as polarizing as this subject is, I HIGHLY doubt that 1200 people accurately reflect the attitude of the ENTIRE country. but think what you want, free country and all.

      • Wogi@lemmy.world
        link
        fedilink
        arrow-up
        2
        ·
        1 year ago

        1200 is a little light but not outside the norm. 1500-3000 participants is bog standard for a population the size of the US. Much more is really overkill and a waste of money. You really are popping a blood vessel over the last decimal point of accuracy.

        • FontMasterFlex@lemmy.world
          link
          fedilink
          arrow-up
          1
          arrow-down
          1
          ·
          1 year ago

          jfc, no one is ‘popping a blood vessel’. if I call 3000 people in the US South, I’m going to get a far different opinion on something than if i call 3000 people in the Pac NW.