ABSTRACT
This chapter introduces hypothesis testing and illustrated how to apply the concepts when testing a hypothesis about a population proportion. The accuracy of the p-value from the one sample t test depends on how well the sampling distribution of the mean can be described with the t distribution. The theoretical development of the t distribution as applied to testing the value of a single mean was based on the assumption that the population from which the sample is derived is normally distributed on whatever is being measured. The permutation test conceptualizes chance in a paired measurement design such as this as a random ordering of the paired evaluations. Rejection of the null hypothesis of random ordering of measurements justifies a conclusion that there is some systematic ordering, which we can use to justify a claim in this case that “chance” is a very plausible explanation for the obtained difference between the means.
