Discussion of statistical methods for determining purity of citrus juice.

It is difficult to develop statistical methods to determine the purity of fruit juice because of country-to-country and environmental differences among fruit. A test of an unknown sample of citrus juice can err in 2 ways: Pure juice can mistakenly be rejected as adulterated, or adulterated juice can be accepted as pure. Many of the statistical procedures proposed may misclassify a high proportion of samples of pure juice as adulterated. It is necessary to develop a statistical test that will only rarely reject samples of pure juice, and will have a good chance of rejecting adulterated juice.