The formula for the confidence interval is

z_{/2} | n |

Thus, the error in the confidence interval is

Error = | z_{/2} | n |

where we have s = 0.5, = 0.05, /2 = 0.025, z_{/2} = 1.96, s = 0.5. We are also required to come up with an error of no more than 0.1, and we are asked to estimate n. To do this, substitute these values in the formula and solve for n:

0.1 = | 1.96 | n |

This gives

- n = 1.96 0.5/0.1 = 9.8.

- n = 9.8

Thus, to be absolutely shure the confidence interval is no wider than 0.1, you should round this up to 97 people (since 96 people would give an error slightly larger than 0.1). Thus, you should have at least 97 people in the survey.

Close this window to return to the tutorial.

Just close this window to return to the exercises.

Lost the exercises to which this window is attached? Press here.