ABSTRACT
This chapter deals with concepts for testing inferences about variance, in particular, the sampling distributions underlying tests. It illustrates use of the test of a single variance. The chapter also deals with several inferential tests of variance. Although the sampling distribution of the mean is a normal distribution and the sampling distribution of a proportion is either a normal or chi-square distribution. A researcher at the esteemed Ivy-Covered University is interested in determining whether the population variance in intelligence at the university is different from the norm-developed hypothesized variance of 225. Several tests have traditionally been used to test for the equality of independent variance. The Brown-Forsythe procedure is a variation of Levene’s test. The Brown-Forsy is recommended for leptokurtic distributions, as it is robust to nonnormality and provides adequate Type-I error protection and excellent power.
