1

I want to know the fraction of bananas which are ripe at some moment in time.

If I have 100 bananas, and 25 of them are ripe, then my rate is 25%. But if I want error bars on this so I can generalize – would I use N=100 and say "25 ± 10 bananas are ripe" (25±10%) or would I use N=25 and say "25 ± 5 bananas are ripe" (25±5%)?

tomr_stargazer
  • 111
  • 1
    You have a set total. This cannot be a Poisson. I’d suggest using a binomial instead and constructing a confidence interval on the proportion. – Daeyoung Mar 25 '22 at 05:21
  • 1
    @DaeyoungLim oh that's excellent! Thank you! I've taken this approach: https://en.wikipedia.org/wiki/Binomial_proportion_confidence_interval#Normal_approximation_interval – tomr_stargazer Mar 25 '22 at 14:24

0 Answers0