> the main problem was that the 95% confidence intervals for these estimates were between 2 and 520 people.
> the main problem was that the 95% confidence intervals for these estimates were between 2 and 520 people.