Stephen Kent Campbell - Flaws and Fallacies in Statistical Thinking

When a person loves something he hates to see it abused. Such is my feeling toward statistics. Hardly a day goes by that I fail to see the subject I love put to use in faulty and misleading ways in newspapers, magazines, books, speeches, and, especially, advertising.

Curiously, little of the statistical literature is aimed at helping the nonstatistician recognize such abuses. Statisticians, understandably, tend to write for other statisticians. Still, one of lifes rare truisms is that most people in the world are not statisticians and dont aspire to become statisticians. Nevertheless, it is an ususual person who isnt frequently confronted with the need to evaluate, if only very informally, statistical information when making decisions about a variety of subjects ranging from which political candidate to vote for to which brand of detergent to buy. Professional statisticians dont need help along such lines; keen critical judgment about statistical information is, for most, a natural by-product of their professional training. But who is to help the statistical layman if it is not the statistical professional?

This book was written with two purposes in mindthe first, less important, purpose being that of getting something off my chest. For many years I have been distressed by the frequency with which (1) relatively simple statistical tools such as percents, graphs, and averages are misused and (2) faulty conclusions are drawn from, perhaps, flawless data in our news media simply because the purveyors of the information dont know any better. Moreover, I have been annoyedindeed, made damned madby the frequency with which bogus statistical evidence is used intentionally by some unconscionable people to sell their products or pet ideas to others. Writing this book has been good therapy for me.

The second, more important, purpose of this book is loosely related to the first. I have long felt that the university student who is likely to take only one or two courses in statistics and the ordinary citizen who maybe lacks any classroom exposure whatever to the subject could benefit from a nontechnical book written with a view to helping him increase his ability to judge the quality of statistical evidence and, in turn, to make better-informed decisions about many facets of everyday life. This book, therefore, has been written both as a supplemental reading text for the student taking his first course in statistics and as a self-help guide for the nonstudent who feels the need to evaluate statistical evidence more judiciously than he is presently capable.

The sequence of topics covered is in rough accord with that of many beginning. textbooks in statistics. The terminology used and the manner in which the subjects are treated are based on the assumption that the reader has had little or no prior exposure to the subject of statistics and has studied precious little formal mathematics.

Many people have helped me with the preparation of this bookmore than I can possibly thank individually. I must, however, single out three statistics professors for special thanks. Professors Richard B. Ellis of Northern Essex Community College and Richard E. Lund of Montana State University both reviewed early drafts of the first five chapters and offered many incisive criticisms and imaginative suggestions for improving upon my proposed project. My colleague and friend, Professor Paul R. Merry of the University of Denver, went over the final manuscript with a thoroughness that is so characteristic of everything he does and offered numerous constructive suggestions.

I would also like to thank my statistics students at the University of Denver who gathered hundreds of examples of statistical fallacies from which many of those appearing in the following pages were selected. Also deserving of thanks are several authors, editors, and advertisers who granted me permission to quote from copyrighted material even though they must have suspected that I intended to be more critical than complimentary.

Last, but certainly not least, I wish to thank my wife Judy who not only relinquished without complaint many hours of time with me which were rightfully hers but who also did more than her fair share to create an atmosphere within which the work could proceed with a bare minimum of discord or distraction.

Needless to say, any blame for errors, omissions, or bad manners should be directed at me.

STEPHEN K. CAMPBELL