Business Analysis with Microsoft® Excel, Second Edition

Understanding Confidence Intervals

In many situations, it is either too expensive or completely impossible to acquire all the information that would allow you to make a decision with 100% confidence.

Suppose you manufacture cars, and want to test how well the bumpers perform in a 15 mph crash. To conclude, “Less than $1,000 damage” with complete confidence, you'd have to test all your cars. And then you'd have nothing left to sell.

In this case (and in less drastic ones), the best that you can do is to obtain a sample of the information. With the sample in hand, you can calculate a statistic that you hope is a close approximation of the value you would calculate if you had access to the entire information set. You can measure the “closeness” ...