Business Analysis Using Regression : A Casebook

Preface Statistics is seldom the most eagerly anticipated course of a business student. It typically has the reputation ofbeing aboring, complicated, and confusing mix of mathematical formulas and computers. Our goal in writing this casebook and the companion volume (Basic Business Statistics) was to change that impression by showing how statistics gives insights and answers interesting business questions. Rather than dwell on underlying formulas, we show how to use statistics to answer questions. Each case study begins with a business question and concludes with an answer. Formulas appear only as needed to address the questions, and we focus on the insights into the problem provided by the mathematics. The mathematics serves a purpose. The material is organized into 12 "classes" of related case studies that develop a single, key idea of statistics. The analysis of data using statistics is seldom very straightforward, and each analysis has many nuances. Part ofthe appeal ofstatistics is this richness, this blending of substantive theories and mathematics. For a newcomer, however, this blend is too rich and they are easily overwhelmed and unable to sort out the important ideas from nuances. Although later cases in these notes suggest this complexity, we do not begin that way. Each class has one main idea, something big like standard error. We begin a class by discussing an application chosen to motivate this key concept, and introduce the necessary terminology. Providing an introduction to modern data analysis techniques, these casebooks can be used as the primary or secondary text for elementary business statistics courses. Statistics has the reputation of being a boring, complicated, and confusing mix of mathematical formulas accompanied by computers used to do something. This casebook and its companion volume Business Analysis Usuing Regression change that impression by showing how statistics gives insights and ansswers interesting business questions.

The material is organized into classes of related case studies that develop a single key idea of statistics. The authors begin by discussing an application that motivates the key idea. Students are then shown how to analyze a data set. The emphasis of the analysis is to answer important business questions with statistics rather than talk about statistics.

Basic Business Statistics introduces ideas often not emphasized in elementary texts such as issues of robustness, the use of transformations to simplify problems, sampling bias, confounding, kernel density, quality control, and scatterplot matrices. Business Analysis with Regression includes a discussion of scatterplot smoothing, prediction intervals for new observations, collinearity, logistic regression, nonlinear models, and multiple comparisons in regression.

The text includes directions for data analyses with JMP and an appendix with Minitab commands. Professors Dean P. Foster, Robert A. Stine, and Richard P. Waterman are members of the Department of Statistics at the Wharton School, University of Pennsylvania.

"...includes a discussion of scatterplot smoothing, prediction intervals for new observations, collinearity, logistic regression, nonlinear models, and multiple comparisons in regression."

"The rapid emergence of strategic collaborations as alternatives to the usual go-it-alone entrepreneurial ventures is evident everywvhere - from the growing collaborative efforts of such muultinationals as IBM, AT&T and Siemens, to Japan's continuing use of alliances to help maintain their enviable economic dominance." "Equally evident is the failure of many companies to make these collaborations work, and the fear that such alliances suggest corporate weakness and lead to an ultimate lack of control." "In his remarkably detailed and lucid account of this major development, Robert Porter Lynch dispels the myths and fears surrounding alliances and provides you with time-tested, practical techniques and tips that enable you to create powerful linkages which secure long-term strategic goals." "Offering his own counsel plus ideas from many of the world's best strategic thinkers, including Peter Drucker, Kenichi Ohmae and Harvard's Michael Porter, along with dozens of expert managers who contributed their best advice in strategizing, negotiating, and structuring alliances, Business Alliances Guide reveals the unique quality of collaborations - why they operate by a different set of rules, what advantages and drawbacks they bring, what goes into their success, what can cause them to self-destruct - and how you can ultimately benefit from them." "Recognizing that alliances are built systematically - not hastily thrown together - the Business Alliances Guide provides an architectural model of an effective collaboration - the Alliance Development Pyramid - which, when followed, will help you avoid costly pitfalls and maximize the chances of successful implementation." "Moreover, Business Alliances Guide alerts you to the "six deadly sins" that have destroyed at least 80% of all failed alliances. It tips you off to the seven early warning signs when an alliance is going awry and what you can do to set things right. And it illustrates, by highlighting the unfortunate plight of companies whose alliance efforts failed, the strategies you can employ to avoid the same fate." "Enhanced by numerous graphs, models, charts, checklists, and point-by-point summaries, as well as boxed-off tips and warnings that underscore especially important advice, the Business Alliances Guide gives you a new and exciting foothold in today's global economy, helping you fully tap the speed, synergy, agility, and relative safety of strategic collaborations."--BOOK JACKET Front Matter....Pages i-xvii Fitting Equations to Data....Pages 1-38 Assumptions in Regression Modeling....Pages 39-83 Prediction and confidence intervals in regression....Pages 85-104 Multiple regression....Pages 105-132 Collinearity....Pages 133-155 Modeling Categorical Factors with Two Levels....Pages 157-176 Modeling Categorical Factors with Two or More Levels....Pages 177-198 Summary Regression Case....Pages 199-227 Comparing Many Mean Values....Pages 229-250 Analysis of Variance with two Factors....Pages 251-271 Modeling a Categorical Response....Pages 273-298 Modeling Time Series....Pages 299-332 Back Matter....Pages 333-348