Ziegel, E.R., Technometrics, May 2002, Volume
44, Issue 2, page 200.
Copyright American Statistical Association, May 2002
Statistics for Lawyers (2nd ed.), by Michael O. FINKELSTEIN and Bruce LEVIN, New York: Springer-Verlag, 2001, ISBN 0-587-95007-9, xxx + 617 pp., $79.95.
Statistics for legal applications is a new genre that is becoming increasingly more commonplace. This is the first book on statistics for lawyers that has come out in a second edition. Statistics books for lawyers have been getting ever better. In this issue I give glowing praise to the fine new book by Good (2001). When it appeared, the first edition of this book was my pick as the best available book of its kind. It was also the first book to fully integrate statistical methodology with its use in litigation in the courts. In the review of the first edition (Ziegel 1992), I wrote: "I strongly recommend this book for any statisticians who must help lawyers understand statistics or who are consulting with lawyers to support their efforts with statistical assistance. This book has been carefully prepared to provide cogent and succinct information about a broad spectrum of statistical methods applicable to jurisprudence. There is complete linkage to actual cases in which statistics has played a role in the judicial process." My description of the first edition was fairly complete by my editorial report standards.
The most impressive change in the book can be attributed to the publisher, who has included the book in its relatively new series on Statistics for Science and Public Policy. First published as an inexpensive softcover edition, the book is now a beautiful hardcover volume at twice the price of the first edition, although it is certainly not overpriced. The most significant improvement is the layout. The statistical content for each of the chapters creates the sections, and the cases provide subsections. In the first edition, everything was a section. The authors highlight the new topics for the cases in the second edition: "DNA evidence, epidemiologic studies in toxic substance litigation, statistical models for adjusting census counts, and vote dilution cases" (p. vii). There is an entirely new chapter on statistics in epidemiology. Other new content includes Monte Carlo methods, additional probability distributions (geometric and exponential), and several new regression methods. The very long regression chapter was split, and the second chapter, on the more complex methods, now covers time series models and locally weighted regression.
More comprehensive than the book by Good (2001) and authored jointly by a lawyer and a statistician, this book continues to be a wonderful resource for the use of statistics in the courts. It would be wonderful to know what my father, now deceased, a trial judge for 36 years in county and state courts in Ohio, would think of this book and his need for some comprehension of statistical methods in understanding testimony and rendering decisions. Maybe he would even be more impressed now with his son who did not choose to become a lawyer.
REFERENCESFrom Hutton, J.L. (2002), Biometrics, Volume 58, page 605:
...There are 21 examples, from vascular surgery and commodity trades to torture in the Philippines, in the chapter on statistical inference for two proportions. A sensible discussion of measures of the differences between proportions precedes theory, which includes Fisher and chi-squared test, maximum likelihood estimation, power and confidence intervals. Later material builds on these standard concepts of estimators in covering topics such as linear, probit, logistic and Poisson regression, times series, survival analysis, Monte Carlo methods and epidemiology.From Haigh, J. (2002), The Statistician [Journal of the Royal Statistical Society, Series D], Volume 51, Part 4, page 577:
Although the fields of law and statistics have interacted for many years, recent developments in natural science (e.g. DNA profiling) and social science (e.g. equal opportunities, discrimination and possible deterrence effects of sentencing) have made us more aware of their links. This new edition is an expanded version of the book that was first published in 1990 and forms a happy collaboration between a practising lawyer who also teaches and a distinguished statistician with considerable experience of testifying in court. It will also be useful to statistics teachers in fields away from law, both for the quality of the exposition and for the variety of statistical issues covered. Few statistics books are as rich in real data as this.From Amazon.com:
From the reviews of the second edition:
"The dangers of misusing statistics are illustrated by the case of
People vs. Collins, in which the product of the probabilities of several
characteristics described by eyewitnesses was presented as convincing
evidence against the defendant. This book analyses this and numerous
other interesting examples to show the relevance of statistics to forensic
science, medicine, commerce and the law. Each of the statistical concepts
and procedures is presented with examples that include exercises based on
actual data. The book is quite readable."
Dr. Henry Roberts, Australian Journal of Forensic Sciences, Vol. 35,
2003)
"This book is designed to introduce law students, law teachers,
practitioners, and judges to the basic ideas of mathematical probability
and statistics as they are applied in the law. The cases and case studies
reflect a broad variety of legal subjects, including antidiscrimination,
mass torts, taxation, school finance, identification evidence and the
death penalty. Significant new developments, reflected in this edition,
include, for example, DNA evidence, epidemiologic studies in toxic
substance litigation, statistical models for adjusting census counts, and
vote-dilution cases."
Edward M. Psyadlo, Zentralblatt MATH, Vol. 983, 2002)
"***** One of the best books on statistics and the law, January 22,
2008
A reviewer for a statistics journal called Phil Good's 'Applying
Statistics in the Courtroom' the best book ever written on statistics and
the law. Given the fine books by Finkelstein, by Gastwirth and the
deGroot, Feinberg and Kadane book, I think that was a gross overstatement.
Phil does however get away from legalese and tries to present key issues.
He does a nice job.
Finkelstein is a lawyer who with Herbert Robbins helped define what
statistical evidence should be. He is well educated in statistics and his
first edition was a classic. This book maintains the good features of the
first book and provide a nice update with modern advances particularly in
genetics. It is an introductory statistics text for lawyers with little or
no statistical background and it teaches them the methods utilizing legal
cases as examples.
I was very much impressed with the authors' analysis of the Florida
vote in the 2000 Presidential election that Finkelstein presented in a
talk at the Joint Statistical Meetings in New York in August 2002. I heard
the talk and discovered that this meticulous and interesting analysis is
covered in the book, section 4.5.3 "Election 2000: Who won Florida?" This
detail is typical of the nice interplay between statistical methodology
and important legal questions. It is just one example of the gems in this
book!
The chapters are 1. Descriptive Statistics, 2. How to Count, 3. Elements
of Probability, 4. Some Probability Distributions, 5. Statistical
Inference for Two Proportions, 6. Comparing Multiple Proportions, 7.
Comparing Means, 8. Combining Evidence Across Independent Strata, 9.
Sampling Issues, 10. Epidemiology, 11. Survival Analysis, 12.
Nonparametric Methods, 13. Regression Methods, 14. More Complex Regression
Models.
With many interesting and famous cases as examples this book is valuable
to statisticians like me as well as to attorneys."
Michael R. Chernick 'statman31147' (Holland PA)
Here are some excerpts from reviews of the first edition of
Statistics for LawyersFrom The Institute of Mathematical Statistics Bulletin, Vol. 19, pp. 708-709, 1990:
"A series of several dozen case studies (e.g., Dalkon shield, automobile emission, death penalty, draft lotteries) arranged by relevant statistical methodology (e.g., descriptive statistics, counting, statistical inference, sampling, regression), interspersed with brief synopses of the relevant theoretical issues and commentary on actual court decisions. Numerous exercises require both calculation (full solutions at the end of the volume) and interpretation. Of interest both to future lawyers and to future statisticians; excellent source of ?relevant? examples."
Lynn Arthur Steen, St. Olaf College
From the Michigan Law Review, Vol. 89, pp.1520?1544, May 1991:
"I conclude that any prospective or practicing attorney who works through Statistics for Lawyers will indeed emerge with ?a fuller appreciation of the standards for analyzing data and making inferences? (p. ix). ?Statistics for Lawyers succeeds admirably in its goal of introducing the ideas and techniques of statistics that have the most application to the courtroom and to the formulation of legal doctrine. Despite its pedestrian title, it is not a routine statistics text with legal examples tossed in. The selection of topics and examples, as well as the exposition of statistics and law, is erudite, informed, and even entertaining. With a strong and steady hand, Statistics for Lawyers opens the tools chest of the professional statistician, permitting students of the law to peer within."
David H. Kaye, Arizona State University College of Law
From Journal of the American Statistical Association, Vol. 86, pp. 544-546, June 1991:
"For its modest price it is a ?best buy? entrée to a smorgasbord of interesting legal issues and data sets. When supplemented by several thoughtful opinions, a truly exciting educational experience awaits both students and teacher. The main attraction of the book is that, whether or not one agrees with the suggested analysis or with the end result of the legal process, Statistics for Lawyers stimulates the reader to think. What more can one ask of scholarly text?"
Joseph L. Gastwirth, George Washington University
From Choice USA, January, 1991:
"Finkelstein and Levin?s book will greatly benefit both lawyers who struggle with the problem of understanding statistical analyses and statisticians who struggle with the problem of understanding the approaches used in legal settings. The book consists of a long, well-organized series of case studies, interlaced with sections explaining pertinent concepts from mathematical statistics and short problem sets that allow readers to check their mastery of the concepts. The mathematical sections will be moderately challenging to readers who have taken one or two semesters of undergraduate statistics. Although the authors claim that their aim is to introduce lawyers and prospective lawyers to methods of statistical analysis used in legal disputes, this book has a real place in any undergraduate statistics library, and should also be required reading for statisticians who may be called upon to present statistical analyses in legal disputes."
F. Giesbrecht, North Carolina State University
From Technometrics, Vol. 34, pp. 122-123, 1992:
"I strongly recommend this book for any statisticians who must help lawyers understand statistics or who are consulting with lawyers to support their efforts through statistical assistance. This book has been carefully prepared to provide cogent and succinct information about a broad spectrum of statistical methods applicable to jurisprudence. There is complete linkage to actual cases in which statistics has played a role in the judicial process. "
Eric R. Ziegel, Amoco Corporation
Back to Bruce Levin's home page.