ABSTRACT
In the selection task, the evaluation of the hypothesis tested by the tester can result in a demonstration of either its complete truth or its complete falsity, according to the standard of proposition logic. There is no intermediate conclusion possible, even if such a conclusion might be highly intuitive to an everyday reasoner. In most common hypothesis-testing situations, observations provide uncertain information about the hypotheses; they provide some support for a hypothesis, but no decisive support. When we test someone's personality by asking him the question, “do you like parties?”, we cannot definitely conclude that he is socially awkward if he answers “no”, but we do feel that we have gained some information about his character. The uncertainty assumption is an assumption made by statistical theories on testing (see Chapter 3). In these theories, tests themselves can be described in terms of the amount of support or counter-evidence they can give a hypothesis. The truth values of hypotheses can vary continuously from 100% certainty about their truth to 100% certainty about their falsity.
