A Test of Normality: Especially Against Symmetric Alternatives presents a practical statistical method for evaluating normality, with a focus on symmetric alternatives.
This edition explains how to use order statistics and regression ideas to test whether data come from a normal distribution.
The book guides readers through exact and approximate testing approaches, discusses power considerations, and shows how to compare results with other standard tests. It emphasizes practical steps, including how to handle multiple replications and how to interpret test statistics in familiar terms like the beta and F distributions.
- How to frame normality testing as a regression problem using order statistics
- Ways to compute test statistics and assess their distribution under normality
- Practical guidance on power and sample size for robust decisions
- Comparisons with other common normality tests to judge effectiveness
Ideal for readers of introductory and applied statistics who want a clear, implementation-focused guide to assessing normality and understanding the behavior of these tests in symmetric contexts.