FDA Caught Between Opposing Protesters

Statistical Tests and Confidence Intervals

Two (more) statistical principles should be a part of every newspaper reader’s protection kit. The first deals with the trade-offs inherent in any probabilistic situation. Suppose, for example, I hypothesize that at least 20 percent of the people in a certain region are blond but that, upon observing one thousand people in representative parts of the region, I note only eighty blonds among them. Using probability theory I calculate that, given my assumption, the likelihood of this result is well below 5 percent, a commonly used “level of significance.” Therefore I reject my hypothesis that 20 percent of the people in the region are blond.

There are two sorts of ...

Start Free Trial

No credit card required