In the preface, the author of this popular text (now in its 6^{th} edition) states that this book is meant to be a nontheoretical introduction to the elements of statistics, utilizing just basic algebra. The goal is to explain basic concepts intuitively with abundant examples and applications in diverse fields which utilize statistics. In this regard, the author has done an admirable job.
As with the previous editions, the author covers the basic topics that one would expect in an elementary text. There is undoubtedly more than enough material for a three credit semester course. A quick internet search uncovered several instructors who utilized the 5^{th} edition, and their course outlines appear to cover the first 10 of the 14 chapters. Topics covered in these ten chapters are as follows: frequency distributions and graphs, measures of central tendency and variation, introductory probability, discrete probability distributions (mainly binomial), the normal distribution, confidence intervals and sample size, hypothesis testing, testing the difference between two means and two proportions, and correlation and regression. There are an enormous number of exercises, and instructors who have used the 5^{th} edition may be interested to know the number of exercises have been increased from about 1800 to 2100.
There are numerous positive features of this book. There is a constant attempt on the part of the author to show how the material is relevant to real world applications, e.g., research studies, media articles, business, medical, social science, etc. Real data is constantly used in the examples and exercises. We all know that students often find statistics a difficult subject, and the literal “stepbystep” approach obviously is very appealing to the beginning student of statistics. Instructors are aware that students may have forgotten some of their basic algebra; fortunately, there is an algebra review section in the appendix.
This reviewer feels that the author was generally correct in marking certain items as “optional”. For example, for the discrete probability distributions, the limited material on multinomial, poisson and hypergeometric distributions were designated as optional; full material of course was provided for the binomial distribution, which is more widely used in statistical practice. Multiple regression was discussed in Chapter 10 as an optional topic and Analysis of Variance (ANOVA) was placed near the end of the text (Chapter 12). This also seems reasonable, since multiple regression and ANOVA often serve as the basis of a second course in statistics.
Another positive feature of the book is the decision by the author to let the instructor determine how much technology he or she wishes to utilize for the course. The author provides detailed writeups on performing the statistical tests, whether one uses a graphics calculator, EXCEL, or MINITAB. The latter is a package specifically designed for statistical analysis, and would give the student a feel as to how the professional statistician does his or her “number crunching”.
This reviewer was pleasantly surprised to see some very nice useful practical topics and applications in the book, often not seen in other books. For example, when discussing percentiles, the author provided a percentile chart of weights of girls by age (the kind of tool a pediatrician would use). Also, a “rule of thumb” was discussed, i.e., a standard deviation can sometimes be estimated by calculating 25% of the range. Very gratifying was the wealth of material on sample size determinations for confidence intervals of the mean and proportions. After all, a basic question that a client (who was once probably an elementary statistic
This reviewer feels there are, nevertheless, a few relatively minor shortcomings to this book. For one thing, the book will probably retail for well over $100 (the 5^{th} edition was about $115 at Barnes and Noble). In addition, the student might be interested in purchasing the study guide, the student solution manual, and/or one of three specially written manuals which describes the software used for the course (i.e., graphics calculator manual, the Minitab 14 Student Version manual, or EXCEL manual). These “addons” could probably make the total cost of books for the course in excess of $200, which seems like an expensive outlay for just an introductory course.
Another potential problem with the text is that material on nonparametric statistics (techniques which do not assume normally distributed data) appears near the end of the text (Chapter 13). It is not unusual to see this topic relegated to the end (and consequently often not covered due to lack of time). The fact is that much of real data is not normally distributed; consequently, student
A final point of criticism is that some algebrabased statistical texts will in fact provide a few elementary mathematical proofs which utilize just basic algebra. In particular, the basic properties of summation (sigma) notation could be developed, and a few simple proofs may be presented. For example, the simple proof which derives the computational formula for the sample variance from the definitional formula is often provided in such texts. Although the author did state in the preface that no proofs are provided, what harm would be done in providing (possibly as an optional topic) a few such algebraicbased proofs?
Despite these minor shortcomings, this book clearly dispels the popular contention that statistics is a dry and dull topic; the colorful nature of the material (literally and figuratively) of the varied applications discussed in the text should be very appealing and friendly to both instructor and student.
Martin Feuerman is a biostatistician in the Department of Academic Affairs at WinthropUniversity Hospital in Mineola, Long Island,New York. He previously served in a similar capacity at the University of Medicine and Dentistry of New Jersey. He has also taught elementary statistics at the Bronx campus of Mercy College.
