Types of Hypothesis and definition

Answer 1

In formal design and analysis of experiments there are but two types of hypotheses: null and alternative. And one might argue there really is only one because when the null is properly defined, the alternative is automatically properly defined.

The null hypothesis is a testable statement of conjecture. The purpose of the null hypothesis is to set the measurable goal for the experiment that follows to show that the null is not false. If the results of the experiment do not show that then the alternative hypothesis is by definition not false.

Simple Example:

Null: It's raining outside.

Alt: It's NOT raining outside. NOTE: The NOT reverses the logic of the null.

The experiment...walk outside. The test...if I get wet, the Null is not false. If I don't get wet, the alternative is not false.

NOTE: I must have an experiment to test the hypothesis. Without a test it's not a valid hypothesis.