ABSTRACT
The second area of statistical inference is that of hypothesis testing. A statistical hypothesis is simply a statement concerning the probability distribution of a random variable. Before the hypothesis is formulated, it is necessary to choose a probability model for the population. There are two types of general hypotheses: simple and composite. A simple hypothesis is one which states that the data in question are represented by a distribution that is a specific member of a particular family of distributions. For example, the hypothesis that µ = 6, CI2 = 1.7 for a normal population uniquely specifies a particular normal distribution. If the value of the variance is unknown and we stated the hypothesis as µ 6, we would have a composite hypothesis. This is because the hypothesis states that the distribution involved can be any member of the family determined by the parameters µ = 6 and
> 0. Once the hypothesis has been stated, appropriate statistical procedures
are used to determine whether it is an acceptable conjecture or an unaccept able one. In general, we cannot prove that a hypothesis is absolutely true or false. If the information furnished by the data supports the hypothesis, we do not reject it. If the data do not support the hypothesis, we reject it.
