Types of Hypotheses

On the basis of structure statistical hypotheses can be of two types:
Simple Hypothesis: If a statistical hypothesis specifies the population distribution completely, then it is called Simple Hypothesis. From the above stated example we can say that hypothesis “H: ? = 17” is a simple hypothesis provided the variance s2 is known, since it is specifying the population distribution completely.

Composite Hypothesis: If the statistical hypothesis does not specify the population completely it is called Composite Hypothesis. On the other hand “H: ? = 17” can be considered as a composite hypothesis since it does not specify the population distribution completely. This classification was on the structural basis.
Another classification is based on the seriousness of rejection i.e. which one makes a more serious error when gets rejected. On the basis of this we classify them as Null Hypothesis and Alternative Hypothesis.