Statistical Inference to Compare Parameters from Two Populations

Our main tools for statistical inference are confidence intervals and hypothesis testing. Although it may seem that statistical inference is only slightly more complicated, there are several things that require consideration: whether samples are independent or dependent; whether the standard deviations are known; if unknown, whether the standard deviations can be assumed to be equal or cannot be assumed to be equal. This chapter addresses these questions. It shows how to conduct inference on two population means using confidence intervals when variances are known and are unknown. It should come as no surprise that on occasion, the comparison of two population proportions is needed. Also, just like when comparing population means, it is convenient to compare two proportions by their difference. The chapter also features practice problems designed to ensure that readers understand the concepts and can apply them using real data.