these were all sampled from 1 company in amsterdam. the differences could be explained by company culture, or local culture, or whatever. more work needed.

[–] Telodzrum@lemmy.world 29 points 4 months ago (1 children)

n=40 isn’t actually bad for generalized conclusions, given a reasonable spread in the results. Your second point is a much stronger argument. The sample is entirely non-representative.

[–] ramjambamalam@lemmy.ca 9 points 4 months ago (1 children)

IIRC from stats n=32 is generally considered the minimum to be considered representative for a random sample (and this is not a random sample outside of the company in Amsterdam 🙄).

[–] elbarto777@lemmy.world 1 points 4 months ago (2 children)

I don't think you're disagreeing with the parent poster...

[–] luciferofastora@lemmy.zip 7 points 4 months ago (1 children)

Not every reply is a disagreement. It's a common assumption that we're not always aware of, because it's true often enough.

[–] elbarto777@lemmy.world 3 points 4 months ago (1 children)

I don't think we're disagreeing.

[–] luciferofastora@lemmy.zip 3 points 4 months ago (1 children)

Neither do I. In fact, I'd be so bold as to say we agree!

Have a nice day :)

[–] elbarto777@lemmy.world 3 points 4 months ago

You too!

[–] ramjambamalam@lemmy.ca 5 points 4 months ago

That's correct.

[–] jwt@programming.dev 25 points 4 months ago

That's very concrete language you're using there. Are you perchance an introvert? We could make it n = 41 and add a dash more selection bias to boot!

[–] SineSwiper@discuss.tchncs.de 2 points 4 months ago

Shitty sample sizes are the majority of "research" nowadays. It's sad how hard it is to find any even in the triple digits.

[–] pyre@lemmy.world 2 points 4 months ago

anything with personality types i already assume is junk. might as well use their zodiac sign.