CHAPTER VIII - MORE TESTS OF HYPOTHESES

Abstract

This chapter discusses how to test hypotheses about the mean of a normal population when the population variance is not known and to test hypotheses about the mean of populations that are not normal. Statistical tests of hypotheses about the mean of a distribution that is not normal and whose variance may be unknown are only approximate tests. One must have a large sample if the population from which one is sampling is not normal. If the population is normal, then a small sample is sufficient and one uses either the normal table or the t-table to test a hypothesis about the mean of the population, depending upon whether the population variance is known or unknown. Occasionally, one may have two binomial populations and wants to test whether the proportion of successes is the same for both or, more generally, that the difference between the two proportions of successes is a specified number.