Conversation

I have bad sense for low sample sizes; is 50 people from a subpopulation meaningful to get averages from, like how hard should I try to scrape together more data points? I know how confident to be in it but not like, what usefulness other ppl take from it

Aella

@Aella_Girl

My completely intuitive sense based on just exposure to data collection over time is that like... 150-175 starts to be a good number that makes me feel relaxed

6:58 PM · Aug 28, 2022

Replying to

I think that sounds good. When making inferences, larger samples make for more confidence in your results. How other people "take usefulness from it" probably depends more on how you present your data and articulate yourself.

Replying to

Intuition-driven statistical sample sizing. No t-test required! Those pesky T’s, math is hard

Replying to

I remember Tim Urban posting a screenshot of a poll he did at 2% of the eventual responses and they were essentially identical. He said he was amazed at how representative that early sample was.

Replying to

50 is OK for looking at basic mean differences among groups 200+ for looking at regression analysis Some of your samples are so large that you need to avoid over-interpreting small effects that have no practical importance even if the effect is statistically reliable.