Statistical methods /

The objective of Statistical Methods, Second Edition, is to provide students with a working introduction to statistical methods. Courses using this book are normally taken by advanced undergraduate statistics students and gradúate students from various disciplines. Statistical Methods is an upper-level requirement for undergraduate degrees in disciplines emphasizing quantitative skills, or a requirement for gradúate degrees in disciplines where statistics is an important research tool. This book is intended to be used for this type of course. The material in this book provides an overview of a wide range of applications and normally requires two semesters, although a limited knowledge of statistical methods is provided in the first semester. Many students will continué with several-additional courses in specialized statistics applications. Traditionally, textbooks used for statistical methods courses have emphasized plugging numbers into formulas, with computer usage as an afterthought. This approach has led to much mind-numbing drill, which obscures the real issues. The increased usage of computers and availability of comprehensive statistical software packages wouid seem to imply that statistical methods shouid now be taught in terms of implementing such software. This approach is likely to make the computer appear as a black box into which one pours data files and automatically receives the correct answers. However, a computer does not know whether it is doing the correct analysis and is capable of a beautifully annotated execution of an incorrect analysis. Aiso a computer cannot interpret results and write a report. Guiding Principies This text provides a reasonable compromise between these two extremes. Our guiding principies are as follows: • No mathematics beyond algebra is required. However, mathematically oriented students may still find the material in this book challenging, especially if they are aiso exposed to courses in statistical theory. • Formulas are presented primarily to show the how and why of a particular statistical analysis. For that reason, there is a minimal number of exercises that plug numbers into formulas. • The topics in this book are organized in broad categories to facilitate the choice of the best-performance methodology for a specific task and there is considerable cross-referencing to facilítate making this choice. • All examples containing real data are worked to a logical conclusión, including interpretation of results. Where computer printouts are used, the results are discussed and explained. In general, the emphasis is on conclusions rather than mechanics. • Throughout the book we stress that certain assumptions about the data must be fulfilled in order for the statistical analyses to be valid, and we emphasize that although the assumptions are often fulfilled, they shouid be routinely checked. New to This Edition - Friendiier exposition makes concepts clearer to students without weakening the statistical rigor of the material. - New, greater emphasis on graphics help students to visualize and understand ideas. - Examples of contemporary topics, such as analysis ofmeans, are included at appropriate points in the text. - Exercises or portions of exercises are identified when material is covered from specific sections, allowing students to practice the methods without having to wait until a complete chapter is covered. - Examples and exercises contain both contemporary data and references to additional data on the Internet or in other published works.