Practical Data Analysis with JMP®

Conditions for Inference

At the end of the prior chapter, we studied the matched pairs design. In this chapter we take up the comparison of samples drawn independently from two different populations, and learn to estimate the differences between population means and variances. The procedures discussed here assume that samples are randomized (and therefore representative) and are a comparatively small proportion of their respective populations.

We'll first focus on the centers of the two populations, and assume that measurements are independent both within and across samples. As we did in the last chapter, we begin with cases in which the populations are either approximately normal or in which the Central Limit Theorem applies. Then we'll see ...