FROM THE BOOKSHELF

ANDREW RUPPEL, Feature Editor, McIntire School of Commerce,
University of Virginia

Textbooks using spreadsheets

by Vijay S. Desai, McIntire School of Commerce, University of Virginia

Spreadsheets have been one of the work-horse applications that drove the PC revolution. Over the years spreadsheets have grown in power, with the latest versions offering features that can help the user perform statistical analysis, solve linear and nonlinear optimization problems, and perform Monte Carlo simulation. Textbooks in the statistics and OR/MS area that fully integrate these spreadsheet features are now appearing in the market place, and those of us who are on the lookout for textbooks for the coming academic year have quite a few choices.

The established textbooks in these areas, with a few notable exceptions, typically have used stand-alone packages (e.g., MINITAB or SPSS for statistics, and LINDO for linear programming) or add-ins (e.g., What'sBest! or FAST QM for OR/MS). Textbooks that integrate spreadsheet features are an important development because, unlike stand-alone packages, spreadsheets are widely available, and a large number of managers are already familiar with modeling in spreadsheets. Also, at present, users do not have to learn a new package for every new tool. These apparent advantages have prompted some observers to predict that the inclusion of statistics and OR/MS features in spreadsheets will lead to the increased usage of these concepts and a resurgence in these fields.

Below, I describe some of the books that cover statistics or OR/MS using spreadsheets. By no means is this a complete list of books using spreadsheets, nor are the reviews complete. I simply describe some of the features I found interesting.

Perhaps one of the first statistics books to integrate spreadsheets into all the topics, this book uses Lotus 1-2-3 macros, called Statistical Applications Macros (SAM), for probabilities, descriptive statistics, inferences, and regression analysis. While some familiarity with spreadsheets is expected, the book includes self-contained instructions in spreadsheets integrated with the teaching of statistics. Although SAM will work in the Windows environment, they were originally written for the DOS version of Lotus 1-2-3 and hence do not take advantage of the latest features available in the Windows version. Another distinguishing feature of the book is numerous well-written cases focused on real life applications of statistics. The book can be used in undergraduate as well as MBA classes, and for those of us who feel uncomfortable using full-blown Harvard cases in undergraduate classes, this is an invaluable source. The emphasis of the book is on developing the students' ability to understand and address real problems in a sensible and useful way, and in its exposition the book stays away from using too many formulae or lengthy calculations.

This book starts with an introduction to Excel and how to manage and print files in Excel. Subsequent chapters describe how to use all the statistical features available in Excel. The last two chapters cover how to construct formulae for statistical measures and methods not readily available in Excel, such as the Durbin-Watson statistic, autocorrelation, and autoregression. Coverage of statistics per se is minimal as the book is intended to be a how-to supplement for a regular statistics text. As there are a number of ways for performing the same analysis in Excel, the book illustrates a variety of them, starting with the simplest. The author suggests that the data sets for the examples are intentionally small and can be quickly entered by the user, or copied via anonymous ftp by the user, or requested from the publisher at a nominal price.

In comparison to the previous book, this book can serve either as a supplementary text or as the main text for an introductory statistics course. In addition, this book comes with two very useful features. First, it has statistical add-ins that, among other things, generate indicator variables, boxplots, histograms, multiple histograms, normal probability plots, and X-bar and range charts. I found that these add-ins merge seamlessly with the other Excel features and do not require much additional effort to use. Second, the authors have created interactive worksheets, called instructional templates, which allow the students to actually work with the basic concepts of statistics rather than reading about them. Instructional templates are included for important concepts such as the central limit theorem, confidence intervals, and hypothesis tests, and have good pedagogical value. Also, the instructor's version comes with 50 fairly interesting real-life data sets, such as data on 392 car models on variables such as miles per gallon, number of cylinders, engine displacement, and country of origin.

This book is one of a four-part series on Managerial Decision Analysis by the two authors. The other three parts cover decision making under certainty, decision making under uncertainty, and risk management, respectively. Essentially, all four books are a collection of Harvard Business School teaching notes and cases prepared by the two authors. For this volume, the teaching notes and cases are supplemented with Excel spreadsheets and two utilities, one for regression and another for displaying cumulative distributions. Some of the advantages of using the regression utility are that variables can be named, noncontiguous columns can be used, observations can be excluded, and more than sixteen independent variables can be used. These features are not available in the regression tool provided by Excel. Also, the authors claim, and I agree, that their regression output has less clutter and hence is easier to interpret than the one provided by Excel.

One of the first management science books to integrate spreadsheets into all topics, this book uses DOS-based spreadsheets and the What'sBest! solver software. A distinguishing feature of this book is the use of influence charts to guide the modeling process and to explain models. Topics covered include linear, nonlinear, and integer programming, decision analysis, and simulation. Since the emphasis of the book is on modeling and managerial usefulness, several cases are provided. Also, to emphasize the importance of models over techniques, several cases are repeated with different solution approaches illustrated. The author suggests that the book can be used for undergraduate as well as graduate classes.

This book uses Excel 5.0 for Windows for illustrations throughout the textbook. Also, OR/MS success stories reported in magazines and journals, some as recent as 1993, are included. Consistent with the spreadsheet orientation, the focus is on model formulation and implementation, and hence details of algorithmic procedures such as the simplex method are kept to a minimum. In addition to covering the traditional topics such as linear, integer, and nonlinear programming, networks, simulation, queuing theory, and project management, this book also includes a good coverage of topics like regression analysis, time-series analysis, and discriminant analysis. These topics are not typically part of a MS/OR course. Monte Carlo simulation is covered using the @RISK add-in, and the coverage of decision analysis is aided by the use of a shareware called TreePlan. @RISK is also used for simulating problems in the project management and decision analysis areas. Given the broad scope of topic coverage, instructors may find this book especially appealing. The addition of a chapter on data analysis would make this a perfect book for a two-semester sequence on introductory statistics and management science.

Vijay S. Desai is an Assistant Professor in the MIS/QM area at the McIntire School of Commerce, University of Virginia at Charlottesville where he teaches statistics, management science, and operations management. He received his Ph.D. and M.B.A. degrees from Indiana University, a M.M.S. in finance from Bombay University, and a B. Tech. in Chemical Engineering from the Indian Institute of Technology at Bombay. Dr. Desai's research interests include the application of dynamic optimization and artificial intelligence techniques in business. He is a member of Decision Sciences Institute, TIMS, ORSA, and POMS. He has published in several journals including the Annals of Operations Research, the European Journal of Operational Research, and Computer and Mathematical Modeling.

Dr. Andrew Ruppel
Monroe Hall
University of Virginia
Charlottesville, VA 22903
Voice mail: 804-924-3867
Fax: 804-924-7074
e-mail: acr2y@virginia.edu