• The cell phone case, which is the first continuing case in Chapter 1, discusses how a bank uses a random sample to estimate its cell phone costs. Using this estimate, the bank decides whether to outsource management of its wireless resources. This case should be particularly motivating to students because it addresses a real problem faced by both students and businesses—unpleasantly high cell phone bills.

• Continuing cases with no need to refer back to previously given computer outputs. Each time a continuing case is revisited, any needed computer output is included with the current case discussion. In addition, whenever possible the background information needed to understand the current analysis is provided, so the student does not need to refer back to previous material.

• Business improvement icons placed in the page margins identify when an important business conclusion has been reached using statistical analysis. Each conclusion is also highlighted for additional emphasis.

• Confidence intervals for and hypothesis tests about a population mean presented by using the s known/s unknown approach. This approach simplifies the choice of z or t-based procedures and is consistent with computerized procedures provided by MINITAB, Excel, and MegaStat®. A t distribution table with up to 100 degrees of freedom is given in Table A.20 of Appendix A. Confidence intervals for and hypothesis tests about the difference between two population means are also presented using the s known/s unknown approach.

• Completely updated end of chapter computer appendices that clearly show how to perform statistical analysis using MINITAB (Version 14), Microsoft Excel 2003, and the latest version of MegaStat®.

• Expanded coverage of sampling in Chapter 1. We now discuss using both a random number table and computer generated random numbers to select a random sample. We also have added an optional section that introduces stratified, cluster, and systematic sampling and discusses the problems of undercoverage, nonresponse, and response bias.

• A substantial number of new, real world data sets in the exercises, particularly in the exercises of Chapter 1 and Chapter 2 (Descriptive Statistics).

• An optional appendix on covariance and correlation. This end of book appendix (Appendix B) can be covered either after covering scatter plots in Chapter 2 or before covering simple linear regression analysis in Chapter 11. Or, it can be omitted entirely without loss of continuity.

• An optional appendix on normal probability plots. This end of book appendix (Appendix D) supplements the normal distribution discussion in Chapter 5.

• A simpler and easier to understand example introducing sampling distributions. This stockreturn example motivates the discussion of the sampling distribution of the sample mean in Chapter 6.

• Increased emphasis on the concept of the margin of error to better motivate the discussion of confidence intervals in Chapter 7.

• A step-by-step hypothesis testing approach that is used in almost all hypothesis testing examples in Chapter 8 (Hypothesis Testing) and Chapter 9 (Statistical Inferences Based on Two Samples). This approach consists of a seven-step procedure that is designed to break hypothesis testing down into small, easy to understand steps and to also clearly show how to use the book’s hypothesis testing summary boxes. Although the seven-step procedure is not formally used after Chapter 9, the students’ familiarity with the steps and summary boxes should enable them to successfully carry out hypothesis tests in later chapters.

• Increased emphasis in Chapter 9 on the “unequal variances” t-based procedure for comparing two population means. This procedure is becoming increasingly popular because it is available in most statistical software packages and is a very accurate approximation that does not require assuming equal population variances.

• A simplified and improved discussion of simple and multiple regression analysis. In simple regression (Chapter 11), we give more concise explanations of the simple linear regression model, least squares, and confidence and prediction intervals. In addition to using improved graphics, the chapter provides the flexibility to cover simple coefficients of determination and correlation (Section 11.6) early or later in the chapter. In multiple regression (Chapter 12), we have refined the innovative, modular organization of the third edition. This will make it easier to selectively cover whatever multiple regression (and model building) topics are desired. We have also simplified the presentation of dummy variables and added a short section on logistic regression. In both the simple and multiple regression chapters, we have improved our explanations and use of MINITAB, Excel and MegaStat® regression outputs. Key outputs are more clearly annotated to help the beginner find needed regression quantities.