4 ms·
> quite low likelihood if the sampling is truly random... If you're assuming their user base is 50/50 male/female, which for many apps is not a valid assumptio
by nrmitchi 5y ago
> quite low likelihood if the sampling is truly random...
If you're assuming their user base is 50/50 male/female, which for many apps is not a valid assumption.
If I remember how to do math correctly, 50/50 gives 5 random users all being female ~4%. And 80/20 split is closer to 40%.
- bjornsing 5y agoTrue. My implicit assumption was 50/50 or predominantly male, but I could be wrong of course. 4% is a pretty low likelihood though. Far below the level that would warrant further exploration in this kind of situation.