ABSTRACT
This chapter considers several situations where it is of interest to compare two samples. It describes two samples of correlated data. These are data that consist of pairs of observations measuring comparable quantities. The chapter examines two independent samples from populations with different variances. Paired comparisons involve pairs of observations on similar variables. The chapter explores the problem of testing whether the variances of two populations are equal. The observations in each sample are independent of each other and independent of all the observations in the other sample. The most commonly used two-sample technique consists of comparing independent samples from two populations with the same variance. The motivation for testing equality of variances is frequently to justify the use of the pooled estimate of the variance. The equal variances test assumes that the data are independent and normal and that the variances are equal.
