Suppose I am trying, by means of stasitical inference, to determine the number of bicycles owned by left-handers in Gotham Village.

First I take a random sample of all Gotham Village Residents and determine, at a 90% confidence level, that there are 360 +/- 20 left handers in the village.

Then I take a random sample of left handers and determine, again at 90% confidence level, that each left hander owns 1.9 +/- 0.2 bicycles.

My question is: if I am to assert that there are between 578 and 798 bicycles owned by left handers in Gotham Village, what is the relevant confidence level?

NB: 578 is the lower bound at 90% x the lower bound of bicycles. Ie 340 x 1.7, and so on for 798.

First I take a random sample of all Gotham Village Residents and determine, at a 90% confidence level, that there are 360 +/- 20 left handers in the village.

Then I take a random sample of left handers and determine, again at 90% confidence level, that each left hander owns 1.9 +/- 0.2 bicycles.

My question is: if I am to assert that there are between 578 and 798 bicycles owned by left handers in Gotham Village, what is the relevant confidence level?

NB: 578 is the lower bound at 90% x the lower bound of bicycles. Ie 340 x 1.7, and so on for 798.

Last edited: