xv

Statistics is the science of data. The Practice of Statistics for Business and Economics (PSBE) is an introduction to statistics for students of business and economics based on this principle. We present methods of basic statistics in a way that emphasizes working with data and mastering statistical reasoning. PSBE is elementary in mathematical level but conceptually rich in statistical ideas. After completing a course based on our text, we would like students to be able to think objectively about conclusions drawn from data and use statistical methods in their own work.

In PSBE we combine attention to basic statistical concepts with a comprehensive presentation of the elementary statistical methods that students will find useful in their work. We believe that you will enjoy using PSBE for several reasons:

Themes of This Book

Look at your data is a consistent theme in PSBE. Rushing to inference—often automated by software—without first exploring the data is the most common source of statistical error that we see in working with users from many fields. A second theme is that where the data come from matters. When we do statistical inference, we are acting as if the data come from a properly randomized sample or experimental design. A basic understanding of these designs helps students grasp how inference works. The distinction between observational and experimental data helps students understand the truth of the mantra that “association does not imply causation.” Moreover, managers need to understand the use of sample surveys for market research and customer satisfaction and the use of statistically designed experiments for product and process development and improvement.

xvi

Another strand that runs through PSBE is that data lead to decisions in a specific setting. A calculation or graph or “reject H0” is not the conclusion of an exercise in statistics. We encourage students to state a conclusion in the specific problem context, and we hope that you will require them to do so.

Finally, we think that a first course in any discipline should focus on the essentials. We have not tried to write an encyclopedia, but to equip students to use statistics (and learn more statistics as needed) by presenting the major concepts and most-used tools of the discipline. Longer lists of procedures “covered” tend to reduce student understanding and ability to use any procedures to deal with real problems.

What’s New in the Fourth Edition

Content and Style

PSBE adapts to the business and economics statistics setting the approach to introductory instruction that was inaugurated and proved successful in the best-selling general statistics texts Introduction to the Practice of Statistics (eighth edition, Freeman 2014). PSBE features use of real data in examples and exercises and emphasizes statistical thinking as well as mastery of techniques. As the continuing revolution in computing automates most tiresome details, an emphasis on statistical concepts and on insight from data becomes both more practical for students and teachers and more important for users who must supply what is not automated.

Chapters 1 and 2 present the methods and unifying ideas of data analysis. Students appreciate the usefulness of data analysis, and realizing they can actually do it relieves a bit of their anxiety about statistics. We hope that they will grow accustomed to examining data and will continue to do so even when formal inference to answer a specific question is the ultimate goal. Note in particular that Chapter 2 gives an extended treatment of correlation and regression as descriptive tools, with attention to issues such as influential observations and the dangers posed by lurking variables. These ideas and tools have wider scope than an emphasis on inference (Chapters 10 and 11) allows. We think that a full discussion of data analysis for both one and several variables before students meet inference in these settings both reflects statistical practice and is pedagogically helpful.

xviii

Teachers will notice some nonstandard ideas in these chapters, particularly regarding the Normal distributions—we capitalize “Normal” to avoid suggesting that these distributions are “normal” in the usual sense of the word. We introduce density curves and Normal distributions in Chapter 1 as models for the overall pattern of some sets of data. Only later (Chapter 4) do we see that the same tools can describe probability distributions. Although unusual, this presentation reflects the historical origin of Normal distributions and also helps break up the mass of probability that is so often a barrier that students fail to surmount.

We use the notation rather than for Normal distributions. The traditional notation is, in fact, indefensible other than as inherited tradition. The standard deviation, not the variance, is the natural measure of scale in Normal distributions, visible on the density curve, used in standardization, and so on. We want students to think in terms of mean and standard deviation, so we talk in these terms.

In Chapter 3, we discuss random sampling and randomized comparative experiments. The exposition pays attention to practical difficulties, such as nonresponse in sample surveys, that can greatly reduce the value of data. We think that an understanding of such broader issues is particularly important for managers who must use data but do not themselves produce data. Discussion of statistics in practice alongside more technical material is part of our emphasis on data leading to practical decisions. We include a section on data ethics, a topic of increasing importance for business managers. Chapters 4 and 5 then present probability. We have chosen an unusual approach: Chapter 4 contains only the probability material that is needed to understand statistical inference, and this material is presented quite informally. The sections on probability models, general probability rules, and random variables have been reorganized so that they are now self-contained in this chapter. Chapter 5 now focuses on distributions of counts and proportions with new material on checking binomial and Poisson assumptions. It also concludes with a section titled “Toward Statistical Inference,” which introduces the concepts of parameters and statistics, sampling distributions, and bias and precision. This section provides a nice lead in to Chapter 6, which provides the reasoning of inference.

The remaining chapters present statistical inference, still encouraging students to ask where the data come from and to look at the data rather than quickly choosing a statistical test from an Excel menu. Chapter 6, which describes the reasoning of inference, is the cornerstone. Chapters 7 and 8 discuss one-sample and two-sample procedures for means and proportions, respectively, which almost any first course will cover. We take the opportunity in these core “statistical practice” chapters to discuss practical aspects of inference in the context of specific examples. Chapters 9, 10, and 11 present selected and more advanced topics in inference: two-way tables and simple and multiple regression. Chapters 12, 13, and 14 present additional advanced topics in inference: quality control, time series forecasting, and one-way analysis of variance.

Instructors who wish to customize a single-semester course or to add a second semester will find a wide choice of additional topics in the Companion Chapters that extend PSBE. These chapters are:

Chapter 15 Two-Way Analysis of Variance

Chapter 16 Nonparametric Tests

Chapter 17 Logistic Regression

Companion Chapters can be found on the book’s website:

www.macmillanhighered.com/psbe4e.

xix

Accessible Technology

Any mention of the current state of statistical practice reminds us that quick, cheap, and easy computation has changed the field. Procedures such as our recommended two-sample and logistic regression depend on software. Even the mantra “look at your data” depends—in practice—on software because making multiple plots by hand is too tedious when quick decisions are required. What is more, automating calculations and graphs increases students’ ability to complete problems, reduces their frustration, and helps them concentrate on ideas and problem recognition rather than mechanics.

We therefore strongly recommend that a course based on PSBE be accompanied by software of your choice. Instructors will find using software easier because all data sets for PSBE can be found in several common formats on the website www.macmillanhighered.com/psbe4e.

The Microsoft Excel spreadsheet is by far the most common program used for statistical analysis in business. Our displays of output, therefore, emphasize Excel, though output from several other programs also appears. PSBE is not tied to specific software. Even so, one of our emphases is that a student who has mastered the basics of, say, regression can interpret and use regression output from almost any software.

We are well aware that Excel lacks many advanced statistical procedures. More seriously, Excel’s statistical procedures have been found to be inaccurate, and they lack adequate warnings for users when they encounter data for which they may give incorrect answers. There is good reason for people whose profession requires continual use of statistical analysis to avoid Excel. But there are also good, practical reasons why managers whose life is not statistical prefer a program that they regularly use for other purposes. Excel appears to be adequate for simpler analyses of the kind that occur most often in business applications.

Some statistical work, both in practice and in PSBE, can be done with a calculator rather than software. Students should have at least a “two-variable statistics” calculator with functions for correlation and the least-squares regression line as well as for the mean and standard deviation. Graphing calculators offer considerably more capability. Because students have calculators, the text doesn’t discuss “computing formulas” for the sample standard deviation or the least-squares regression line.

Technology can be used to assist learning statistics as well as doing statistics. The design of good software for learning is often quite different from that of software for doing. We want to call particular attention to the set of statistical applets available on the PSBE website: www.macmillanhighered.com/psbe4e. These interactive graphical programs are by far the most effective way to help students grasp the sensitivity of correlation and regression to outliers, the idea of a confidence interval, the way ANOVA responds to both within-group and among-group variation, and many other statistical fundamentals. Exercises using these applets appear throughout the text, marked by a distinctive icon. We urge you to assign some of these, and we suggest that if your classroom is suitably equipped, the applets are very helpful tools for classroom presentation as well.

Carefully Structured Pedagogy

Few students find statistics easy. An emphasis on real data and real problems helps maintain motivation, and there is no substitute for clear writing. Beginning with data analysis builds confidence and gives students a chance to become familiar with your chosen software before the statistical content becomes intimidating. We have adopted several structural devices to aid students. Major settings that drive the exposition are presented as cases with more background information than other examples. (But we avoid the temptation to give so much information that the case obscures the statistics.) A distinctive icon ties together examples and exercises based on a case.

xx

The exercises are structured with particular care. Short “Apply Your Knowledge” sections pose straightforward problems immediately after each major new idea. These give students stopping points (in itself a great help to beginners) and also tell them that “you should be able to do these things right now.” Most numbered sections in the text end with a substantial set of exercises, and more appear as review exercises at the end of each chapter.

Acknowledgments

We are grateful to the many colleagues and students who have provided helpful comments about PSBE, as well as those who have provided feedback about Introduction to the Practice of Statistics. They have contributed to improving PSBE as well. In particular, we would like to thank the following colleagues who, as reviewers and authors of supplements, offered specific comments on PSBE, Fourth Edition:

Ala Abdelbaki, University of Virginia
Diane Bean, Kirkwood Community College
Tadd Colver, Purdue University
Bryan Crissinger, University of Delaware
Douglas Antola Crowe, Bradley University
John Daniel Draper, The Ohio State University
Anne Drougas, Dominican University
Gary Evans, Purdue University
Homi Fatemi, Santa Clara University
Mark A. Gebert, University of Kentucky
Kim Gilbert, University of Georgia
Matt Gnagey, Weber State University
Deborah J. Gougeon, University of Scranton
Betsy Greenberg, University of Texas at Austin
Susan Herring, Sonoma State University
Paul Holmes, University of Georgia
Patricia Humphrey, Georgia Southern University
Ronald Jorgensen, Milwaukee School of Engineering
Leigh Lawton, University of St. Thomas
James Manley, Towson University
Lee McClain, Western Washington University
Glenn Miller, Pace University
Carolyn H. Monroe, Baylor University
Hayley Nathan, University of Wisconsin–Milwaukee
Joseph Nolan, Northern Kentucky University
Karah Osterberg, Northern Illinois University
Charles J. Parker, Wayne State College
Hilde E. Patron Boenheim, University of West Georgia
Cathy D. Poliak, University of Houston
Michael Racer, University of Memphis
Terri Rizzo, Lakehead University
Stephen Robertson, Southern Methodist University
Deborah Rumsey, The Ohio State University
John Samons, Florida State College at Jacksonville
Bonnie Schroeder, The Ohio State University
Caroline Schruth, University of Washington
Carl Schwarz, Simon Fraser University
Sarah Sellke, Purdue University
Jenny Shook, Pennsylvania State University
Jeffrey Sklar, California Polytechnic State University
Rafael Solis, California State University, Fresno
Weixing Song, Kansas State University
Christa Sorola, Purdue University
Lynne Stokes, Southern Methodist University
Tim Swartz, Simon Fraser University
Elizabeth J. Wark, Worchester State University
Allen L. Webster, Bradley University
Mark Werner, University of Georgia
Blake Whitten, University of Iowa
Yuehua Wu, York University
Yan Yu, University of Cincinnati