The error is 1/2 of the confidence interval. Since we want 99% confidence, we have `z_(alpha/2)=2.58` .

Then the error E is given by`E=2.58sqrt((pq)/n)` where p is the percentage with a given characteristic, and q the percentage without. Here p=.14,q=.86, E=.02 and we want to solve for n:

`.02=2.58sqrt((.14*.86)/n)`

`.007751938=sqrt((.14*.86)/n)`

...

## Unlock

This Answer NowStart your **48-hour free trial** to unlock this answer and thousands more. Enjoy eNotes ad-free and cancel anytime.

Already a member? Log in here.

The error is 1/2 of the confidence interval. Since we want 99% confidence, we have `z_(alpha/2)=2.58` .

Then the error E is given by`E=2.58sqrt((pq)/n)` where p is the percentage with a given characteristic, and q the percentage without. Here p=.14,q=.86, E=.02 and we want to solve for n:

`.02=2.58sqrt((.14*.86)/n)`

`.007751938=sqrt((.14*.86)/n)`

`16641=n/.1204` Square both sides; then reciprocate both sides

`n=2004` rounding to the next integer.

**Thus the minimum sample size is 2004 to yield an error of 2 percent with a 99% confidence.**

**Further Reading**