41 Short Book

Reviews

Vol. 21. No. 3 — December 2001 Editor Dr. A.M. Herzberg

REVIEWS

ANNOTATED READINGS IN THE HISTORY OF translated one from French; and S.L. Lauritzen one from . H.A. David and A.W.F. Edwards. New York: Danish. Five more articles are reproduced in their original Springer-Verlag, 2001, pp. xv + 252, US$69.95/DM151.00. English. Each article is introduced by an essay called “Comments on…”; these comments are informative, Contents: interesting and beautifully written, and contain numerous 1. The introduction of the concept of expectation (Pascal, modern connected references. The production is first class. 1654) H.A. David has used parts from this book “in a short course 2. The first formal test of significance on the , recently, given at Iowa State (Arbuthnott, 1710) University.” The collection is fun to browse. Statistics history 3. Coincidences and the method of inclusion and buffs and browsers should order this book immediately. exclusion (Montmort, 1713; N. Bernoulli, 1713; deMoivre, 1718) University of Wisconsin 4. The determination of the accuracy of observations Madison, U.S.A. N.R. Draper (Gauss, 1816) 5. The introduction of asymptotic relative efficiency (LaPlace, 1818) 6. The logistic growth curve (Verhulst, 1845) THE LADY TASTING TEA. How Statistics Revolutionized 7. Goodness-of-fit statistics (Abbe, 1863) Science in the Twentieth Century. D. Salsburg. New York: 8. The distribution of the sample variance under Freeman, 2001, pp. xi + 340, US$23.95. normality (Helmert, 1876) Contents: 9. The random walk and its fractal limiting form 1. The lady tasting tea (Fisher, Design of Experiments) (Venn, 1888) 2. The skew distributions (Galton and ) 10. Estimating a binomial parameter using the likelihood 3. That dear Mr. Gosset (Student t, and both K. Pearson function (Thiele, 1889) and Fisher) 11. Yule’s paradox (“Simpson’s paradox”) (Yule, 1903) 4. Raking over the muck heap (Fisher at Rothamsted) 12. Beginnings of extreme-value theory 5. “Studies in crop variarion” (Anova and controlled (Bortkiewicz, 1922; von Mieses, 1923) randomisation) 13. The evaluation of tournament outcomes 6. “The hundred year flood” (Tippett and E.J. Gumbel) (Zermelo, 1929) 7. Fisher Triumphant (The logic of Inductive Inference, 14. The origin of confidence limits (Fisher, 1930) 1934) APPENDIX A: English Translations of Papers and Book 8. The dose that kills (Bliss and Probits) Extracts of Historical Interest (Bibliography) 9. The bell shaped curve (Lindeberg, Lévy, Höffding) APPENDIX B: First (?) Occurrence of Common Terms in 10. Testing the goodness of fit (Neyman) Statistics and Probability 11. Hypothesis testing (Neyman and E.S. Pearson) 12. The confidence trick (The AIDS epidemic and Readership: Statistics history enthusiasts confidence sets) The preface tells us that “Interest in the history of 13. The Bayesian heresy (Mosteller and Wallace, de Finetti statistics has grown substantially in recent years...“ How can and Savage) we tell? It is true that the number of historical publications 14. The Mozart of mathematics (Kolmogoroff) has grown, but how many people actually read them and what 15. The worms eye view (F.N. David) do they get out of them? Do you really want to read today a 16. Doing away with parameters (Wilcoxon, Chernoff and translation of a paper that E. Zermelo wrote in German in Savage, Pitman) 1929 about the playing strengths of chess players in a 17. When part is better than the whole (biased sampling; tournament? (The underexplained example in that article Mahalanobis) refers to the famous New York 1924 tournament; however, 18. Does smoking cause cancer? (Doll and Hill vs Fisher) chess players may be puzzled about what we can learn from 19. If you want the best person (Gertrude Cox) the relevant “playing strengths” given, since they mirror the 20. Just a plain Texas farm boy (S.S. Wilks) tournament order exactly.) If at this point in the review you 21. A genius in the family (I.J. Good) are becoming annoyed with the reviewer’s apparent attitude 22. The Picasso of statistics (J.W. Tukey) and are saying impatiently, “Of course we should study this 23. Dealing with contamination (G.E.P. Box) sort of history!”, you will enjoy this book very much. 24. The man who remade industry (W. Edwards Deming) H.A. David translated three articles from the original French, six 25. Advice from the lady in black (S.V. Cunliffe) articles from German, and one from Latin; A.W.F. Edwards 42

26. The march of the martingales (Lévy, Aalen, Andersen, MATHEMATICS OF CHANCE. J. Andel. Chichester, U.K.: Gill, Olshen) Wiley, 2001, pp. xxiii + 235, £39.50. 27. The intent to treat (Peto, Cox, Box, and Rubin) Contents: 28. The computer turns upon itself (Efron) Introduction 29. The idol with feet of clay (Kuhn) 1. Probability Afterword, timeline 2. Random walk Readership: Anyone interested in statistics, especially 3. Principle of reflection statistics students 4. Records 5. Problems that concern waiting The parentheses are reviewer’s additions, indi- 6. Problems that concern optimisation cating topics discussed. 7. Problems on calculating probability A very unusual book, containing many excellent 8. Problems on calculating expectation accounts of statistics in practice. The preface and some 9. Problems on statistical methods other chapters discuss deep issues of statistical philosophy. A 10. The LAD method fair number of amusing errors, e.g. neither of the Guinness 11. Probability in mathematics family’s two peers was Lord Guinness. A most interesting 12. Matrix games read. Readership: All students of probability theory, applied University of Essex statisticians in industry Colchester, U.K. G.A. Barnard This is a compilation of interesting and popular problems concerning mainly probability theory, with some statistics. The material is very accessible, in the most part requiring no more than basic elements of calculus. While THE SUBJECTIVITY OF SCIENTISTS AND THE BAYESIAN there are many old favourites here, there are some novelties APPROACH. S.J. Press and J.M. Tanur. New York: and some problems given a new slant through references Wiley, 2001, pp. x + 274, £57.50. to, for example, Olympiad problems and those which have Contents: appeared in the American Mathematical Monthly. The book is 1. Introduction a translation and modification of the original Czech edition. 2. Selecting the scientists There are some glitches as a result (‘dice’ as singular…), 3. Some well-known stories of extreme subjectivity but most are not crucial. The problems inspire the reader to 4. Stories of famous scientists follow up references and the style is generally very engaging. 5. Subjectivity in science in modern times: The Bayesian This is a very useful supplement to Problems and approach Snapshots from the World of Probability (Blom, Holst and Sandell – Springer-Verlag [1994; Short Book Reviews, APPENDIX: References by Field of Application for Bayesian Vol. 14, p. 22]) and the classic Fifty Challenging Problems Statistical Science in Probability (Mosteller – Addison Wesley, 1965, Dover, Readership: Professional scientists and the general public 1987). with an interest in science, in scientists, and Imperial College of Science, in the methods that scientists use Technology and Medicine This book describes the role of subjectivity and London, U.K. F.H Berkshire preconceptions in science, via a series of vignettes illustrating how famous scientists in history achieved their major advances. Chapter 3 briefly describes how Kepler, ENCYCLOPEDIA OF EPIDEMIOLOGICAL METHODS. Mendel, Millikan, Burt and Mead allowed their precon- M.B. Gail and J. Benichou. Chichester, U.K.: Wiley, 2000, ceptions to influence the data they chose to use on which to pp. xxi + 978, £235.00. base their conclusions (or how they distorted or Contents: manufactured data to match their preconceptions). Chapter From Absolute Risk to Vital Statistics 4 describes the work of Aristotle, Galileo, Harvey, Newton, Lavoisier, Von Humboldt, Faraday, Darwin, Pasteur, Freud, Readership: Epidemiologists, statisticians working in Curie and Einstein. Each of the sections in Chapter 4 is epidemiology divided into a brief historical sketch, an outline of their This volume contains a selection of excellent scientific contribution, a list of their major works, and a articles on many of the concepts, methods and tools that discussion of the role of subjectivity in the work. Most of researchers working in epidemiology require. It is difficult to these people are now regarded as having made a major judge whether the selected topics would satisfy all contribution, but some of them are now regarded as little appetites, as the field is becoming richer and more better than examples of self-deception. It is interesting to diversified. However, when consulting this volume regularly have them all examined from the same perspective, in over the last two months, while investigating new projects which their preconceptions drive their theoretical and supporting students’ dissertations, I have always found developments. comprehensive and clear overviews at hand. As far as the role of bias, preconceptions and All entries are linked to each other by web-style subjectivity is concerned in science, this book is fascinating. cross-referencing and are enriched by up-to-date, but also However, in many of the cases it seems contrived to attach historical, references for more in-depth reading. Most of the it to today’s formal methods of Bayesian inference. contributions are also rich of valuable insights in the topic, Imperial College of Science, although sometimes the “encyclopaedic” style becomes rigid Technology and Medicine and too many classifications and sub-classifications are London, U.K. D.J. Hand offered, for instance with differing listings of types of bias.

The articles are written by experts based in well as some differences. Many of the methodological North America, Europe, Australia, New Zealand and Japan. articles already appeared in the Encyclopedia of Biostatistics Hence there is a wide perspective on several of the topics, as but others have been added to cover specific issues, such as 43

“birth cohort studies” and “cancer registries”, or to introduce 12. General linear hypothesis emerging or expanding fields, such as “genetic 13. Nonparametric statistical inference epidemiology”. Some large topics, like case-control studies, Readership: Students taking postgraduate or final year have several articles from different authors which are often undergraduate courses in mathematics linked-up by short enlightened entries from one of the editors. The book consists of three parts namely: (a) the There is no doubt that this Encyclopedia dedicated to core of the probability; (b) foundations of statistical infer- methods in epidemiology is an invaluable tool for ence; and (c) Chapters 11 to 13 on special topics. There is a researchers involved in medical and public health research as wealth of material and, although the topics are of a it offers an excellent springboard for acquiring and conventional nature, the discussions and special topics are strengthening practical and methodological tools. unique. Most of the presentations give far more depth than one would expect in a text of this type. There are five London School of Hygiene and Tropical Medicine hundred and fifty problems with three hundred and fifty London, U.K. B.L. De Stavola worked examples and one hundred and fifty references included in this comprehensive textbook. The mathematical prerequisites to reading this book are that the students should AN INTRODUCTION TO RANDOMIZED CONTROLLED have had basic courses in linear algebra, set theory and a CLINICAL TRIALS. J.N.S. Matthews. London: Arnold, good background in calculus. This text is for mathematics 2000, pp. xiv + 189, £19.99. specialists and not recommended as a service textbook. Contents: South Bank University 1. What is a randomized controlled trial? London, U.K. S. Starkings 2. Bias 3. How many patients do I need? 4. Methods of allocation STATISTICAL METHODS IN SPATIAL EPIDEMIOLOGY. 5. Assessment, blinding and placebos A.B. Lawson. Chichester, U.K.: Wiley, 2001, pp. x + 277, 6. Analysis of results £55.00. 7. Monitoring accumulating data Contents: 8. Subgroups and multiple outcomes PART I: The Nature of Spatial Epidemiology 9. Protocols and protocol deviations PART II: Important Problems in Spatial Epidemiology 10. Some special designs: Crossover, equivalence and clusters Readership: Graduate statisticians 11. Meta-analyses of clinical trials The author has been a major influence in the Readership: Undergraduate and postgraduate students of development of statistical methods in epidemiology, and statistics has been the principal editor of one recent major volume of papers on disease mapping and spatial epidemiology Over recent decades, randomized controlled (Disease Mapping and Risk Assessment for Public Health clinical trials have become established as the method used to [noted, Short Book Reviews, Vol. 19, p. 52]), and co-author of assess new treatments if claims of the efficacy of a an introductory text on disease mapping (An Introductory treatment are to find widespread acceptance. This book Guide to Disease Mapping, with F. Williams). The former of provides an introduction to the statistical methodology that these possibly represents a more advanced text, with the underpins the randomized controlled trial. Administrative emphasis on more sophisticated modelling and technical aspects of running a trial receive little emphasis in the text detail, whereas the latter contains some of the material but there are many excellent books on clinical trial included here presented at a somewhat more basic level. In methodology that the interested reader may consult. Trials comparison, this book aims to provide an intermediate level with binary outcomes are given less prominence in the text introduction to a subject that has recently attracted much than might be expected from their prevalence in medical interest in both public health and applied statistics. It is practice, and survival analysis is completely omitted. organized into two sections, the first a general introduction Sections in the text that contain more advanced material are to the field of spatial epidemiology, comprising five chapters clearly identified with an asterisk so that such material might detailing basic applications, modelling approaches and be omitted on a first reading – thus allowing use of the text statistical formulations, and the second comprising six within courses to be tailored to the needs of the student chapters on specific classic problems of particular public audience. The text assumes no underlying medical health importance or statistical interest. The content of Part I is background, is well written and easy to read. quite general, whereas the second is more statistically explicit. CEFAS Lowesoft Laboratory Finally, five appendices give technical details of the statistical Lowesoft, U.K. C.M. O’Brien and computational approaches used throughout the book. There is an extensive bibliography, and web links where relevant data and software may be obtained. The computa- AN INTRODUCTION TO PROBABILITY AND STATISTICS, tional requirements associated with inference for spatial 2nd edition. V.K. Rohatgi and A.K.M.E. Saleh. New York: epidemiological models are often demanding. Much reliance is Wiley, 2001, pp. xv + 716, £67.95. placed on Markov chain Monte Carlo (MCMC) methods, and especially as packages such as BUGS are available with Contents: specially written spatial analysis modules (GeoBUGS is soon 1. Probability to be available from the Imperial College developers: 2. Random variables and their probability distributions [email protected]). Little discussion is given in this text to 3. Moments and generating functions practical or theoretical MCMC issues, but the implication 4. Multiple random variables given here, that the models described are implementable in 5. Some special distributions practice, is definitely a realistic assessment. 6. Limit theorems This book represents a good general introduction to 7. Sample moments and their distributions spatial epidemiology, with material presented at an 8. Parametric point estimation intermediate level that is possibly most suitable as a 9. Neyman-Pearson theory of testing of hypothesis postgraduate-level text for statistically trained researchers. 10. Some further results of hypothesis testing Overall, it may be a little advanced for non-statisticians, but in 11. Confidence estimation 44

Part I at least, where formal statistical content is fairly 2. Basic statistics minimal, it is still accessible to the non-specialist statistician. 3. Prediction and interpolation Part II is aimed more at the spatial epidemiologist, and 4. Characterizing spatial processes: The covariance and represents many of the standard (and some apparently more variogram contentious) views of statistical modelling in the field. I can 5. Estimating the variogram therefore recommend the book to such a mixed audience. 6. Modelling the variogram 7. Spectral analysis Imperial College of Science, 8. Local estimation or prediction: Kriging Technology and Medicine 9. Cross-correlation, coregionalization and cokriging London, U.K. D. Stephens 10. Disjunctive kriging Readership: Environmental scientists, graduate students of HANDBOOK OF STATISTICAL GENETICS. D.J. Balding, spatial variation M. Bishop and C. Cannings (Eds.). Chichester, U.K.: Wiley, 2001, pp. xxvi + 863, £175.00. As the authors state, “Geostatistics is not easy”; but this well-written and thorough book must surely make the Contents: learning process easier for environmental scientists with PART 1: Bioinformatics some facility in mathematics – to perhaps just high-school PART 2: Population Genetics (English A-Level) standard for much of the content. The PART 3: Evolutionary Genetics book concentrates on modern techniques based on the PART 4: Genetic Epidemiology variogram and kriging. Most of the detail in these techniques PART 5: Animal and Plant Genetics is illustrated by pertinent graphs, diagrams and illustrations, PART 6: Applications using a mixture of real and simulated data. This book should Readership: This is a comprehensive research resource, enjoy the same success as Statistical Methods in Soil and perhaps aimed predominantly at statist- Land Resource Survey, its predecessor by the same authors. icians and applied probabilists, but also for numerate biologists and geneticists, at post- University of Manchester Institute of graduate level and above Science and Technology Manchester, U.K. P.J. Laycock This magnificent book attempts to catalogue and introduce all aspects of modern statistical genetics in a series of thirty chapters contributed by major researchers in the field. Topics range from the fundamental aspects of evolutionary theory and classical statistical genetics, to BLOCK DESIGNS. Volume 1: A Randomization Approach. modern statistical genetics and bioinformatics, such as S. Kageyama and T. Calinski. New York: Springer-Verlag, protein structure prediction. There is certainly suitable 2000, pp. xi + 313, US$49.95/DM114.00/£35.00. material here for a researcher new to the field to gain a good grounding in statistical genetics (if not genetics), but Contents: also sufficient advanced material to merit the attention of 1. Introduction those already familiar with the field. Specific chapters I 2. Basic terminology and preliminaries found particularly interesting and stimulating were those by 3. General block designs and their statistical properties Solovyev on gene prediction, by Weir on forensics, and by 4. Balance and efficiency: Classification of notions Schork et al on pharmacogenetics. I also found the section 5. Nested block designs and the concept of resolvability on evolutionary genetics and phylogenetics particularly Readership: Mathematical statisticians informative. The structure of the book is excellent, and I found the final general index (covering all chapters) This is the first volume of two. It covers in five invaluable. My only minor criticism of the book as a whole is chapters the questions of randomization in incomplete block that, although in the main chapters give sufficient general designs. detail without unnecessary specifics, some chapters omitted This highly mathematical discussion is appropriate information (or relevant references) that would be required in reading for mathematical statisticians who are interested in order to begin an analysis from scratch. However, overall, I the intricacies of design of experiments. It will not be of can thoroughly recommend it. much help to the practitioner who just wants to know how to analyze a balanced incomplete design. Imperial College of Science, It is written by two distinguished researchers in the Technology and Medicine mathematical theory of experimental design. The London, U.K. D. Stephens mathematical treatment throughout is thorough and, at times, difficult and intricate. It contains a mass of GEOSTATISTICS FOR ENVIRONMENTAL SCIENTISTS. information, and introduces the principles with a thoughtful R. Webster and M.A. Oliver. Chichester, U.K.: Wiley, 2001, discussion of the simple randomized design and randomized pp. viii + 271, £55.00. complete block designs in Chapter 1. The authors write in their preface that “this Contents: monograph does not pretend to give an exhaustive 1. Introduction exposition of the theory of experimental design within the randomization approach.” The authors are too modest. It comes very close to being exhaustive and, as a result, some of the intended audience will find it hard and slow to read. The reader will have to wait until the second volume appears to learn about the construction of block designs. University of Texas Austin, U.S.A. P.W.M. John 45

FINITE POPULATION SAMPLING AND INFERENCE. A Readership: Research workers, statisticians, graduate Predictive Approach. R. Valliant, A.H. Dorfman and students R.M. Royall. New York: Wiley, 2000, pp. xvii + 504, £67.95. This is the fourth edition of a popular text whose Contents: third edition appeared eight years ago [Short Book Reviews, 1. Introduction to prediction theory Vol. 13, p. 20]. In this edition, the number of authors has 2. Prediction theory under the general linear model increased by two, the number of pages by about seventy 3. Bias-robustness and the list of references by about three hundred. The text 4. Robustness and efficiency retains its basic original structure and lucidity but has been 5. Variance estimation extensively revised. Chapter 2 is now much more directly 6. Stratified populations aimed at pictorial representation of clusters rather than of 7. Models with qualitative auxiliaries general multivariate data, and the material in the next three 8. Clustered populations chapters has been ordered in a more logical fashion. 9. Robust variance estimation in two-stage cluster Additions include more examples in each chapter, a greater sampling technical background to many of the methods, and 10. Alternative variance estimation methods description of techniques developed in the 1990s. The 11. Special topics and open questions Appendix is much more comprehensive and detailed than before, and over one hundred and twenty of the references are APPENDIX A: Some Basic Tools to work that has appeared since 1992. These are all distinct APPENDIX B: Datasets enhancements, and this continues to be an excellent APPENDIX C: S-PLUS Functions general introduction to an important and expanding topic. Readership: Survey samplers, researchers in survey University of Exeter methods, theoretical and practical Exeter, U.K. W.J. Krzanowski statisticians This text brings together many years of research into the development of the model-based (predictive) APPLIED MULTIVARIATE DATA ANALYSIS, 2nd edition. approach to inference in sample surveys. The basic B.S. Everitt and G. Dunn. London: Arnold, 2001, philosophy is that survey data can be thought of as pp. x + 342. realizations of random variables and that inference can be based on models constructed to reflect this random Contents: process. In this way, the unobserved population units are 1. Multivariate data and multivariate statistics predicted from the observed data using fitted regression 2. Exploring multivariate data graphically models or general linear models. Throughout the book, the 3. Principal components analysis authors compare the model-based approach with the more 4. Correspondence analysis traditional design-based methods, illustrating many new 5. Multidimensional scaling ideas with extensive numerical examples supported by 6. Cluster analysis simulation studies. The sets of data (available by ftp) for 7. The generalized linear model several large surveys, used in many of the examples and 8. Regression and the analysis of variance exercises, are given in an appendix as are a number of S- 9. Log-linear and logistic models for categorical PLUS routines for performing various sampling and multivariate data estimation tasks. 10. Models for multivariate response variables The methods discussed are wide-ranging, 11. Discrimination, classification and pattern recognition including concepts of robustness (bias robust against model 12. Exploratory factor analysis failure) and jackknife variance estimation, and the theory is 13. Confirmatory factor analysis and covariance structure extensively described using heavy matrix algebra with a models liberal use of generalized inverses. The material is aimed at Readership: Experimental scientists, statisticians, researchers in this new conceptually demanding approach undergraduate and master students to the design and analysis of surveys. The presentation is meticulous and the authors have been careful to explain The intermediate-level text book introduces readers these difficult ideas clearly. There are many questions left to a wide range of multivariate analysis techniques. The unanswered which will be a source of ideas for future second edition includes new sections on correspondence research and development. A highly recommended book analysis, neural networks and random effects for repeated which is an essential read for all research workers in this measures. A casualty of breadth is depth, and omitted are area. topics such as, the detail of how non-metric scaling works, how one might fit random-effects models, and some of the University of Southampton possible problems with hierarchical cluster analysis. Southampton, U.K. P. Prescott Mathematical development is not extensive, and confined to boxes and tables. Discussion sometimes presents opposing CLUSTER ANALYSIS, 4th edition. B.S. Everitt, S. Landau views of the utility of techniques, as for instance in the and M. Leese. London: Arnold/New York: Oxford University summary section on exploratory factor analysis. This book is Press, 2001, pp. ix + 237, £40.00. well-structured, and very well-written, and is bound to be a valuable source of reference for scientists. The many Contents: examples are improved and extended in this edition, and are 1. An introduction to classification and clustering a particular strength. There are over sixty exercises, and 2. Visualizing clusters about a third of these have solutions provided. Anyone who 3. Measurement of proximity reads this book will gain a good overall grasp of the subject. 4. Hierarchical clustering However, in order to carry out analyses, they will need 5. Optimization clustering techniques further knowledge. Appendix A lists main computer package 6. Finite mixture densities as models for cluster analysis web-sites. 7. Miscellaneous clustering 8. Some final comments and guidelines University of Kent APPENDIX: Software for Cluster Analysis Canterbury, U.K. B.J.T. Morgan 46

BAYESIAN STATISTICAL MODELLING. P. Congdon. The word “copula” is immediately intriguing. Chichester, U.K.: Wiley, 2001, pp. x + 531, £45.00. I turned to Chapter 6 which opens on p. 65: “A copula provides a uniform representation of a bivariate distribution F on Contents: the unit square. This result is due to Sklar (1959)... .” 1. Introduction: The Bayesian method, its benefits and Puzzled, I looked in the index. “Copula 65, 112, 162, 184.” implementation Then to the preface: “… copula: a focussed expression of 2. Standard distributions: Updating, inference and dependence between two (or several) random variables, prediction totally stripped of any other characteristics.” Then to p. 2 (not 3. Models for association and classification mentioned in the index!) “… a copula, that is the o.d.f. on the 4. Normal linear regression, general linear models and unit square with uniform marginals.” This book is not easy to log-linear models read. Chapter 2 provides a nice historical record of 5. Ensemble estimates: Hierarchical priors for pooling definitions of dependence and independence; this is strength interesting, but it is marred by scores of minor glitches which 6. Latent variables, mixture analysis and models for non make the reader hesitate and then retrace his steps. response Examples: “Correlation may briefly defined…” (p. 12): 7. Correlated data models “Prompted by Reed’s remarks, a well known American 8. Multilevel models, multivariate analysis and mathematical statistician, H.L. Rietz (1918) was the first, to longitudinal models the best of our knowledge, to study…” (p. 13); “… indication 9. Life table and survival analysis of a greater degree of independence than it actually exists…” 10. Bayesian estimation and model assessment (p. 16); “Reaction to Galton’s discovery in France was very Readership: Researchers, statisticians and others swift.” (p. 23); forty years after Bravais in 1886,…” (p. 24); “C.R. Rao (1893), while reminiscing on the origin and This book is an extremely ambitious and largely development of the correlation coefficient,…” (p. 24); “in a successful attempt to describe the current state of applied most interesting, but perhaps somewhat misleading Bayesian statistical practice. Although some theoretical contribution,…” (p. 28): “The proof use simular arguments as aspects of the Bayesian paradigm are included, the the Hoeffding lemma.” (p. 35); and so on. My complaints are emphasis of the book is strongly on applications, and not about misprints like 1893 (every book has these) but therefore is perhaps close in spirit to the book of Carlin and about the writing style, including heavy misuse of commas, Louis (Bayes and Empirical Bayes Methods for Data which makes smooth reading difficult. Two additional Analysis [Short Book Reviews. Vol. 17, p. 2]). The scope of difficulties occur. The “long dash” which sets off an inserted applications in this book is much broader than Carlin and statement is set as a “short dash” meant to link words as on Louis, however, having numerous examples from the p. 11; and words are sometimes broken at the end of a line biological, medical, physical and social sciences; it is in a ridiculous way, for example: har-m (p. 11); detail-s suitable for researchers in any of these fields who are not (p. 22). necessarily statistically trained. The worked examples include Overall, I applaud the book as a brave effort to set conjugate analysis, classification, (generalized) linear models, out in detail a topic that, although extensive in nature, is hierarchical models, latent variables, mixtures, time series usually given short shrift. The book should serve as a useful and spatial models, multivariate analysis, longitudinal data, reference and/or provide the basis for a challenging seminar and survival analysis. The only major topics that receive course. insufficient attention are Bayesian non- and semi-parametrics (Bayesian flexible modelling), and even these are covered to University of Wisconsin a degree. The computational requirements associated with Madison, U.S.A. N.R. Draper inference for sophisticated Bayesian models are often demanding, and this book leans heavily on Markov chain Monte Carlo (MCMC) methods in general, and on the LIKELIHOOD METHODS IN STATISTICS. T.A. Severini. package WinBUGS in particular. It is a remarkable Oxford University Press, 2000, pp. xi + 380, £55.00. achievement to have carried out such a range of analyses Contents: on such a range of sets of data. 1. Some basic contents I found this book comprehensive and stimulating, 2. Large sample approximations and was thoroughly impressed with both the depth and 3. Likelihood range of the discussions it contains. Although there are 4. First-order asymptotic theory more suitable texts for the study of Bayesian theory, and 5. Higher-order asymptotic theory some of the more difficult or controversial aspects are 6. Asymptotic theory and conditional inference downplayed, I can certainly recommend it as a useful 7. The signed likelihood ratio statistic resource describing the practice of Bayesian statistics. 8. Likelihood functions for a parameter of interest Imperial College of Science, 9. The modified profile likelihood function Technology and Medicine Readership: Research statisticians London, U.K. D. Stephens As the author states in the preface “the emphasis in this book is on the development of statistical methods and a CORRELATION AND DEPENDENCE. D.D. Mari and description of the underlying theory rather than on the S. Kotz. London: Imperial College Press, 2001, statement and proof of precise mathematical results”. pp. xiv + 219, £34.00. Potential readers are assumed to have a good knowledge of graduate-level statistical theory to quite a high standard. Contents: The book then provides an introduction to conditional 1. Notations and definitions likelihoods, Edgeworth expansion and saddlepoint approxi- 2. Correlation and dependence: An introspection mations for regular distributions, plus a summary of recent 3. Concepts of dependence and stochastic ordering results in the field. For statisticians who are not specialists in 4. Copulas likelihood theory, it could perhaps be used as a reference work 5. Farlie-Gumbel-Morgenstern models for dependence for the many results on particular distributions which it quotes – 6. Global versus local dependence between random although the subject index is somewhat limited for such a variables purpose. Alternatively, it could be used as a wide-ranging Readership: Graduates in statistics 47 introductory work for research students intending to HANDBOOKS IN MATHEMATICAL FINANCE. OPTION specialize in some part of likelihood theory. PRICING, INTEREST RATES AND RISK MANAGEMENT. E. Jouini, J. Cvitanić and M. Musiela (Eds.). University of Manchester Institute of Cambridge University Press, 2001, pp. xvi + 669, Science and Technology £80.00/.US$120.00. Manchester, U.K. P.J. Laycock Contents: PROBABILISTIC RISK ANALYSIS: FOUNDATIONS PART I: Option Pricing: Theory and Practice AND METHODS. T. Bedford and R. Cooke. PART II: Interest Rate Modelling Cambridge University Press, 2001, pp. xx + 393, PART III: Risk Management and Hedging £37.50/US$54.95. PART IV: Utility Maximization Contents: Readership: Doctoral students, researchers and PART I: Introduction practitioners who already have a good 1. Probabilistic risk analysis knowledge of mathematical finance PART II: Theoretical Issues and Background 2. What is uncertainty? This book consists of a collection of invited 3. Probabilistic methods papers, by different authors, which review the current state 4. Statistical inference of practice in different areas of mathematical finance. Each 5. Weibull analysis chapter is described as having the structure of a brief PART III: System Analysis and Quantification review of existing results, an outline of more recent results, 6. Fault and event trees and a discussion of outstanding problems, with suggestions 7. Fault trees – analysis for how these might be tackled, but obviously the chapters 8. Dependent failures differ in the amount of weight they put on these different 9. Reliability data bases aspects. The book is not for beginners. The blurb describes 10. Expert opinion it as a ‘handbook’ and ‘comprehensive reference work’, and it 11. Human reliability will certainly be a useful reference work for people 12. Software reliability undertaking research in the area. I have to say also that it PART IV: Uncertainty Modeling and Risk Measurement has been beautifully produced. 13. Decision theory Imperial College of Science, 14. Influence diagrams and beliefs nets Technology and Medicine 15. Project risk management London, U.K. D.J. Hand 16. Probabilistic inversion techniques for uncertainty analysis 17. Uncertainty analysis 18. Risk measurement and regulation Readership: “Numerate readers who have taken a first CONTINUOUS STOCHASTIC CALCULUS WITH university course in probability and statistics, APPLICATIONS TO FINANCE. M. Meyer. Boca Raton, and who are interested in mastering the Florida: Chapman and Hall/CRC, 2001, pp. xvi + 319, conceptual and mathematical foundations of US$89.95/£59.99. probabilistic risk analysis.” Contents: Our technological society has an increasing 1. Martingale theory awareness of risk in all its facets. Brought down to its basic 2. Brownian motion building blocks, risk is about uncertainty in frequency as well 3. Stochastic integration as severity of particular events. Mainly using examples from 4. Application to finance the technology sector (space, nuclear, chemical), the authors build up the various methods and tools needed for a Readership: Those with a background knowledge of quantitative assessment of risk. The presentation is to the measure-theoretic probability and Hilbert point, clear and perfectly high-lighted through well-chosen spaces as well as an interest in stochastic practical examples and exercises. Besides the more calculus and finance technical issues surrounding fault and event-tree analysis, attention is also given to softer issues like expert opinion The purpose of this book is to provide a minimalist, and human reliability. The only glaring omission is the rigorous development of the theory of stochastic calculus with economics of risk analysis; decisions on risk are taken a view to the valuation of derivative securities. The within the context of economic reality and hence the development, from discrete-time martingales through modelling of this part of the equation is important. No doubt stochastic integration with respect to continuous semi- in the years to come we will see a closer link between the more martingales and dynamic trading strategies, is largely reliability-based approach to risk analysis and the more standard, and the author has concentrated on the core economic one currently in use throughout the financial necessary for a discussion of continuous time financial industry. This merger would lead to a truly risk-management models. Though rigorous, the book is perhaps easier to platform. follow than texts devoted exclusively to stochastic calculus I am convinced that Probabilistic Risk Analysis will such as that of Karatzas and Shreve [Short Book Reviews, provide a key input for this convergence effort. I for once Vol. 19, p. 9] without the amusing conversational style of learned a lot from reading this book and wholeheartedly Steele [Short Book Reviews, Vol. 21, p. 9]. The final chapter recommend it for its intended readership. It is the ideal text on applications to finance provides a nice succinct discussion on which to base an undergraduate course. of European options, interest rate swaps, caps and floors. ETH-Zürich University of Waterloo Zürich, Switzerland P.A.L. Embrechts Waterloo, Canada D.L. McLeish 48

TIME SERIES FORECASTING. C. Chatfield. Boca Raton, Jenkins, notably in the use of structural models to explain Florida: Chapman and Hall/CRC, 2001, pp. xii + 267, trend and seasonal effects. A state-space model inevitably US$69.95/£46.99. takes the form of an imperfectly observed vector autoregression, so the Kalman filter figures prominently and Contents: is derived efficiently. To the tasks of Chapter 4 is added that 1. Introduction of parameter estimation. The authors are unaware of the 2. Basics of time-series analysis extensive theory of state-space methods in control and 3. Univariate time-series modelling optimization, which would have indicated a much stronger 4. Univariate forecasting methods development of some of their ideas, but their intuition keeps 5. Multivariate forecasting them on a sound track. 6. A comparative assessment of forecasting methods The non-Gaussian case is largely restricted to the 7. Calculating interval forecasts case of a linear autoregression with a disturbance from the 8. Model uncertainty and forecast accuracy exponential family, and is dealt with by efficient simulation Readership: Researchers and practitioners using fore- methods. Both sections present interesting examples of casting methods in economics, government, applications. operations research, industry, and commerce University of Cambridge Cambridge, U.K. P. Whittle This book is a review of forecasting methods based on the analysis of time-series in their time-domain modelling framework. Not included are methods especially developed for particular fields such as meteorology. THE ESTIMATION AND TRACKING OF FREQUENCY. It is not a text-book; therefore fully worked-out B.G. Quinn and E.J. Hannan. Cambridge University Press, examples are not a significant feature, and it is light on 2001, pp. xi + 266, £37.50/US$59.95. theoretical exposition, with few formulae and equations. Contents: After a very basic summary of time-series modelling 1. Introduction procedures, we meet ARIMA modelling, Box-Jenkins, the 2. Statistical and probabilistic methods Kalman filter, exponential smoothing. We have small 3. The estimation of a fixed frequency sections on non-linear models: TAR (threshold auto- 4. Techniques derived from ARMA modelling regression), regime-switching, GARCH (for modelling 5. Techniques based on phases and autocovariances changes in volatility), neural networks and chaotic series. 6. Estimation using Fourier coefficients This is followed by multivariate methods: transfer function 7. Tracking frequency in low SNR conditions models, vector ARMA models and co-integration. None of these have more than a page or so. The coverage of Readership: Scientists and statisticians working in the prediction intervals, model uncertainty and forecast physical sciences accuracy address important questions not always given It is probably true to say that many statisticians adequate discussion. rarely come across the need to accurately estimate However, the value of this book is not its catalogue sinusoidal frequencies, or determine how many there are, of methods, but the way it draws our attention to recent work, especially in a noisy environment. However, such problems and the sections devoted to comparing the methods and are ubiquitous in areas such as physics and electrical making recommendations as to their merits and application. engineering. This book is a timely study of the statistical In this way it gives insights that might otherwise only be methods which can be employed, and their properties, gained through constant attention to the literature. written by B. Quinn and the late T. Hannan, two of the most Imperial College of Science, active and productive researchers in the field. The text gives Technology and Medicine many of the mathematical details, and perhaps some London, U.K. R. Coleman material would have been better collected in appendices or exercises to improve the readability; also the index is much too brief for such a detailed book. Nevertheless, the book is TIME SERIES ANALYSIS BY STATE SPACE METHODS. a valuable and detailed quality reference source, and the J. Durbin and S.J. Koopman. Oxford University Press, MATLAB code provided will be appreciated by practical 2001, pp. xvii + 253, £35.00. scientists. Contents: Imperial College of Science, 1. Introduction Technology and Medicine 2. Local level model London, U.K. A.T. Walden 3. Linear Gaussian state space models 4. Filtering, smoothing and forecasting 5. Initialisation of filter and smoother PRACTICAL FORECASTING FOR MANAGERS. J.C. Nash 6. Further computational aspects and M.M. Nash. London: Arnold, 2001, pp. xvi + 296, 7. Maximum likelihood estimation £24.99. 8. Bayesian analysis 9. Illustrations of the use of the linear Gaussian model Contents: 10. Non-Gaussian and nonlinear state-space models 1. Why forecasts? 11. Importance sampling 2. Planning the forecasting tasks 12. Analysis from a classical standpoint 3. Measuring how well forecasting goals are met: Part 1 13. Analysis from a Bayesian standpoint 4. Data search, gathering, documentation and 14. Non-Gaussian and nonlinear illustrations management 5. Qualitative forecasting: Long-term Readership: Anyone with a interest in time series models 6. Semi-quantitative methods and analysis 7. Forecasting, risk, and strategy management The first nine chapters – three-quarters of the text 8. Measuring how well forecasting goals are met: Part 2 – are devoted to the linear Gaussian model. The state-space 9. Preliminary data analysis and forecasting formulation is contrasted with the ARIMA models of Box and 10. The preliminary forecast: Concepts and examples 49

11. A strategy for performing forecasting data analysis abundance of empirical control rules they will encounter in later 12. Forecasting trend and season I: Multiple regression life should they choose to become practitioners. 13. Forecasting trend and season II: Smoothing methods Brookfield, U.S.A. C.A. Fung 14. Forecasting trend and season III: Time series decomposition 15. ARIMA and related models for forecasting 16. Using ARIMA models: Other issues and examples SEQUENTIAL MONTE CARLO METHODS IN PRACTICE. 17. Comparing and combining forecasts A. Doucet, N. de Fretas and N. Gordon (Eds.). New York: 18. Variations on the theme of seasonal adjustment Springer-Verlag, pp. xxvii + 581, US$79.95/DM179.00. 19. Mixed and extended models 20. Nonlinear regression modelling Contents: 21. Artificial neural networks 1. An introduction to sequential Monte Carlo methods 22. Building the forecast report 2. Particle filters – A theoretical perspective 3. Interacting particle filtering with discrete observations Readership: Students on business management courses, 4. Sequential Monte Carlo methods for optimal filtering professional managers and administrators 5. Deterministic and stochastic particle filters in state- needing a practical guide to forecasting space models The authors of this text have researched their 6. RESAMPLE-MOVE filtering with cross-model jumps ground thoroughly; I suspect through the wealth of 7. Improvement strategies for Monte Carlo particle filters experience they have gained by teaching forecasting at this 8. Approximating and maximising the likelihood for a level. The book is strongly rooted in practical experience general state-space model and shrewdly expresses the benefits and shortfalls of the 9. Monte Carlo smoothing and self-organising state- statistical techniques described. Minitab and Excel are used space model extensively throughout the book with extra references to 10. Combined parameter and state estimation in tools that can be found on the internet. The book is also simulation-based filtering supported by a complementary website at 11. A theoretical framework for sequential importance www.arnoldpublishers.com/support/nash, which contains sampling with resampling exercises and the sets of data used in the text. The main 12. Improving regularised particle filters emphasis is on straightforward ideas than can be easily 13. Auxiliary variable based particle filters understood and used in the business environment. 14. Improved particle filters and smoothing Advanced topics are covered and explained in a manner 15. Posterior Cramer-Rao bounds for sequential that is clear how these techniques can be put into use. This estimation book would be a useful addition to libraries needing to 16. Statistical models of visual shape and motion provide relevant texts on forecasting. 17. Sequential Monte Carlo methods for neural networks 18. Sequential estimation of signals under model South Bank University uncertainty London, U.K. S. Starkings 19. Particle filters for mobile robot localization 20. Self-organising time series model 21. Sampling in factored dynamic systems 22. In-situ ellipsometry solutions using sequential THE EFFICIENT USE OF QUALITY CONTROL DATA. Monte Carlo K.W. Kemp. Oxford: Clarendon Press, 2001, pp. xi + 260, 23. Manouvering target tracking using a multiple-model £60.00. bootstrap filter Contents: 24. Rao-Blackwellized particle filtering for dynamic PART I: Statistical Concepts Bayesian networks 1. Some aspects of statistical quality control 25. Particles and mixtures for tracking and guidance 2. Small samples: Decisions and consequences 26. Monte Carlo techniques for automated recognition 3. Distributions relevant to process and test control Readership: Those with a research interest in sequential 4. Effective use of sampled data Monte Carlo PART II: Principles and Criteria of Statistical Quality Control 5. Principles and criteria of statistical quality control A comprehensive set of papers. The table of 6. Better control rules contents says it all. PART III: Control Using Cumulative Sums University of Waterloo 7. Really efficient use of test data Waterloo, Canada D.L. McLeish Readership: Mature undergraduates or graduate students in statistics This is an accessible but very intelligent mono- LEVY PROCESSES: THEORY AND APPLICATIONS. graph on the statistics of the quality control. Chapter 1 to 4 can O.E. Barndorff-Nielsen, T. Mikosch and S.I. Resnick (Eds.). nearly stand on their own as a first course in mathematical Boston: Birkhäuser, 2001, pp. x + 415, statistics, while Chapters 5 to 7 develop the foundations of SFr148.00/DM196.00/ÖS1431.00. Shewhart and cusum charts. The monograph is not intended as a handbook for practitioners. Contents: Quality control has always been a small sample 1. A tutorial on Lévy processes problem with attendant risks of misjudging the populations 2. Distributional, pathwise, and structural results being controlled, but the author is right to re-emphasize this 3. Extensions and generalizations of Lévy processes point as the motivation for his monograph: That the use of 4. Applications in physics quality control data must be efficient in the statistical sense so that maximum information can be obtained from any investment in sampling and measurement The monograph is more mathematical than data analytical, but students will be well prepared to negotiate the 50

5. Applications in finance The last third of the book is devoted to the 6. Numerical and statistical aspects concepts and applications of reliability theory. Included therein is a coherent presentation of the main concepts and Readership: Students and researchers in stochastic definitions of reliability, including failure rates, coherent processes (i.e. Lévy processes) and their systems, maintainability and performability. With the aim of applications making the models tractable, several methods for Over the last decade, in numerous fields of computing the key probabilistic quantities that arise in the applications which traditionally used Brownian motion based models are presented and compared. Several specific modelling, non-Brownian models entered the field. One examples including repair and shock models are worked out such class of processes concerns the so-called Lévy along with a brief discussion of model simulations. The processes, i.e. processes with stationary and independent authors have included an extensive list of recent references increments. Lévy based models allow for a more realistic that adds to the overall usefulness of the book. (often heavy-tailed) modelling of the distribution of the University of Washington increments. This volume presents a useful summary of Seattle, U.S.A. R. Pyke some of the recent scientific developments concerning Lévy processes. Both introductory and more advanced articles are included. The interested researcher will get a good overview of ‘where the action is’, whereas students will find STOCHASTIC SPECTRAL THEORY FOR SELF-ADJOINT numerous interesting research topics to work on. The fields FELLER OPERATORS: A Functional Integration Approach. of applications can be deduced from the Contents. It is fair M. Demuth and J.A. van Casteren. Basel, Switzerland: to say that the text is biased towards probabilistic aspects; Birkhaüser, 2000, pp. xii + 463, SFr165.00/ only few papers treat actual data fitting. On the other hand, DM198.00/ÖS1446.00. the field is vast and expanding quickly so that realistically Contents: only a biased snapshot can be expected. I am convinced 1. Basic assumptions of stochastic spectral analysis: that the text will contribute further to making stochastic Free Feller processes models based on general Lévy processes even more 2. Perturbations of free Feller processes popular. I, therefore, take pleasure in recommending this 3. Proof of continuity and symmetry of Feynman-Kac volume to all interested readers. operators ETH-Zürich 4. Resolvent and semigroup differences for Feller Zürich, Switzerland P.A.L. Embrechts operators: Operator norms 5. Hilbert-Schmidt properties of resolvent and semigroup differences SEMI-MARKOV PROCESSES AND RELIABILITY. 6. Trace class properties of self-adjoint Feller operators N. Limnios and G. Oprisan. Boston: Birkhäuser, 2001, 7. Convergence of resolvent differences pp. xii + 222, SFr165.00/DM220.00/ÖS1106.00. 8. Spectral properties of self-adjoint Feller operators Contents: APPENDIX A: Spectral Theory 1. Introduction to stochastic processes and the renewal APPENDIX B: Semigroup Theory process APPENDIX C: Markov Processes, Martingales and Stopping 2. Markov renewal processes Times 3. Semi-Markov processes APPENDIX D: Dirichlet Kernels, Harmonic Measures, 4. Countable state space Markov renewal and semi- Capacities renewal processes APPENDIX E: Dini’s Lemma, Scheffé’s Theorem, 5. Reliability of semi-Markov systems Monotone Class Theorem 6. Examples of reliability modelling Readership: Probabilists APPENDIX A: Measures and Probability APPENDIX B: Laplace-Stieltjes Transform In the classical Schrödinger equation, one has a APPENDIX C: Weak Convergence differential operator – half the Laplacian plus a suitable (Kato) potential, V; the solution is expressed by the Readership: Applied probabilists and reliability researchers Feynman-Kac formula. How do things vary when the Semi-Markov or Markov renewal processes are potential V varies? And how far can one generalize this? ideal for modelling random phenoma that move between a The appropriate generalization is to the theory of set of states like a Markov chain but where the transition Feller processes – in terms of their resolvents, semigroups, times need not be assumed to be either constant or spectral properties etc. – when one compares their exponentially distributed. Because of this great flexibility, potentials. Typically, one is interested in a potential these processes are used for modelling in many subjects. becoming singular on a small set. Various physical contexts The first two-thirds of this monograph provides a reasonably produce problems of this type: Quantum theory, quantum thorough and readable overview of these and their related field theory, atomic or solid-state physics, etc. processes. The essential review of renewal theory that is The authors give a very thorough and detailed included in the first chapter is well done and should be account of results in this area, using a combination of widely useful. Overall, the substantial review materials of probabilistic and analytic methods. Much useful information on the early chapters make this book a recommended the diverse background areas is contained in the reference for probabilistic modellers in many areas of appendices. application. Brunel University Uxbridge, U.K. N.H. Bingham

51

NOTES

BIOMETRIKA: ONE HUNDRED YEARS. D.M. Titterington approximately 50 years old, and it has been at least 15 and D.R. Cox (Eds.). Oxford University Press, 2001, years since such a summary appeared; because much pp. viii + 383. £45.00. progress has been made in recent decades, the time is propitious for this book. The intended audience is primarily From the book jacket: “This volume marks the but not exclusively statisticians working in cancer research; centenary of Biometrika, one of the world’s leading it is hoped that oncologists might benefit as well from academic journals in statistical theory and methodology . A reading this book.” brief history of the journal (D.R. Cox) sets the scene, followed by six specially commissioned articles that review AEONS. THE SEARCH FOR THE BEGINNING OF TIME. the most important contributions made by papers in the journal M. Gorst. London: Fourth Estate, 2001, pp. 314, £14.99. to a number of important areas of statistical activity. These From the book jacket: “When did the world begin? are general theory and methodology (A.C. Davidson), design This question strikes at the root of mankind’s understanding of experiments (A.C. Atkinson and R.A. Bailey), survival of the universe, or of the divine scheme of things. In 1650, analysis (D. Oakes), nonparametrics (P. Hall), sample Bishop James Ussher announced to no one’s enormous surveys (T.M.F. Smith) and time series (H. Tong). surprise that the world began in 4004 BC. His date was so Authoritative but accessible, the reviews provide an impressive widely accepted that it was printed in the margins of the survey of statistical science development during the King James Bible, a close reading of which had allowed twentieth century. Ussher to make the calculation in the first place. But though “The second half of the book consists of reprints of his logic was impeccable, his numbers were wrong. Ussher selected papers that are highlights from the pages of was followed by a sometimes bizarre, always dogged Biometrika. These include: F. Yates (1939) on the design of stream of scientists and philosophers, each determined to agricultural and biological experiments; H.E. Daniels (1944) on calculate the moment when the world began. Edward Lhuyd measures of correlation in sample permutations; took his inspiration from the boulders of his native Wales to E.S. Pearson (1968) on the early correspondence between conclude that the world must be several thousand years W.S. Gosset, R.A. Fisher and K. Pearson; and older than Ussher proposed; Edmund Halley, the W.K. Hastings (1970) on Monte Carlo sampling methods astronomer, suggested that the salinity of the sea might using Markov chains. provide the clue to the age of the earth, meanwhile the “Biometrika: One Hundred years is designed to be Comte de Buffon attempted to replicate the cooling of the accessible, stimulating and useful to all those interested in Earth’s core by heating tiny iron balls and comparing their the development of statistics and in the application of loss of heat. He reckoned the world was 10 million years statistical theory to the problems of science, medicine, old. technology, economics and demography.” “Darwin needed the world to be old enough to STOCHASTIC PROCESSES: THEORY AND METHODS. allow evolution, and by the beginning of the twentieth D.N. Shanbhag and C.R. Rao (Eds.). Amsterdam: century scientists at last started to think in terms of billions Elsevier, 2001, pp. xvii + 967, DFL375.00/€170.17/ of years. In 1998 two independent teams of astronomers US$175.00. probing the heavens with the Hubble space telescope found what they believe is the final piece in the jigsaw – the From the preface: “J. Neyman, one of the pioneers decisive evidence that reveals the age of the universe. With in laying the foundations of modern statistical theory, a margin of error of over a billion years, their result may lack stressed the importance of stochastic processes in a paper the reassuring precision of Ussher’s exact moment of written in 1960 in the following term: “Currently in the period creation, but in a final twist to the tale the discovery unveils of dynamic indeterminism in science, there is hardly a a universe more startling than anyone imagined. Aeons is serious piece of research, if treated realistically, does not the story of science, religion and philosophy, as ever locked involve operations on stochastic processes. … in a tense struggle to define and explain the human “The literature on stochastic processes is very condition.” extensive and is distributed in several books and journals. There is a need to review the different lines of researches ENIGMA. THE BATTLE FOR THE CODE. and developments in stochastic processes and present a H. Sebag-Montefiore. London: Phoenix, 2001, consolidated and comprehensive account for the benefit of pp. xvi + 491, £7.99 paper. [Original 2000] students, research workers, teachers and consultants. … “An effort is made in this volume to cover as many From the book cover: “No episode in the Second branches of stochastic processes as possible. Also to get the World War has captured the modern imagination more balance right, we have retained some chapters with applied strongly than the cracking of the Enigma code by the boffins flavour in this volume. In the planned second volume, we keep at Bletchley Park. the option of including one or two chapters of theoretical “Yet was this really what happened? Without for a nature, assuming that they provide with avenues for future moment belittling the work of Alan Turing and his team of research to specialists in applied areas.” eccentric codebreakers, this book shows the extent to which the breaking of the all-important Naval Enigma code was HANDBOOK OF STATISTICS IN CLINICAL ONCOLOGY. reliant on more traditional forms of cloak and dagger: The J. Crowley (Ed.). New York: Dekker, 2001, pp. xiv + 548, heroic capture of ships and U-boats and their codebooks on US$175.00. the high seas, and the betrayal of his German homeland by From the preface: “This book is a compendium of Enigma Spy, an old-fashioned traitor. Such deeds turned statistical approaches to the problems facing those trying to out to be just as decisive as any cryptographic make progress against cancer. As such, the focus is on breakthrough. Using new material from the archives this cancer clinical trials, although several of the contributions book for the first time tells the full, thrilling story of Enigma: also apply to observational studies, and many chapters The Battle for the Code.” generalize beyond cancer research. This field is 52

ROCKEFELLER AND THE INTERNATIONALIZATION OF From a book review by Roald Hoffmann, Cornell MATHEMATICS BETWEEN TWO WORLD WARS. University: “A world admiring and yet suspicious of science Documents and Studies for the Social History of needs an intelligent way to talk of what we know, and how that Mathematics in the 20th Century. R. Siegmund-Schultze. knowledge is gained. More than moving beyond the Basel, Switzerland: Birkhäuser, 2001, pp. xiii + 341, ‘Science Wars’ the voices of The One Culture? shows us, SFr128.00/ DM170.00/ÖS1241.00. with spirit, how this can be done.” From the book cover: “Philanthropies funded by A GARDNER’S WORKOUT. Training the Mind and the Rockefeller family have been prominent in the social Entertaining the Spirit. M. Gardner. Natick, history of the twentieth century for their involvement in Massachusetts: Peters, 2001, pp. xi + 319, medicine and applied science. This book provides the first US$35.00/£25.00. detailed study of their relatively brief but nonetheless influential foray into the field of mathematics. From the book cover: “Gardner, the Master of “The careers of a generation of pathbreakers in Mathematical Games and Puzzles, has introduced a wide modern mathematics, such as S. Banach, B.L. van der range of readers to the world of recreational mathematics Waerden and André Weil, were decisively affected by their over the past decades, and charmed his way into their becoming fellows of the Rockefeller-funded International hearts with enthusiasm for the subject. Perhaps best known as Education Board in the 1920s. To help promote cooperation the author of Scientific American’s “Mathematical Games” between physics and mathematics Rockefeller funds column for 25 years. Gardner has gained a large following supported the erection of the new Mathematical Institute in among puzzlers, magicians and mathematicians. Göttingen between 1926 and 1929, while the rise of “In the years since Gardner has continued to write probability and mathematical statistics owes much to the articles for academic journals and popular magazines. Forty- creation of the Institute Henri Poincaré in Paris by American one of those pieces, never before published in book form, philanthropy at about the same time.” are collected in this volume. “Truly a treat for Martin Gardner’s many fans, the QUANTUM LEAPS IN THE WRONG DIRECTION. Where articles span a wide range of topics. They include games of Real Science Ends… and Pseudoscience Begins. chance and why a “computer” will always beat a human player, C.M. Wynn and A.W. Wiggins. With cartoons by S. Harris. word ladders and mathematical word play games, tiling Washington, D.C.: Joseph Henry Press, 2001, pp. x + 226, puzzles, magic squares, computer and calculator “magic” US$18.95/Can$25.95. tricks and other mathematical puzzles. Providing the tools to furnish our all-too-sluggish minds with an athletic workout, From the book cover: “Get the straight story on Gardner’s problems foster an agility of the mind as they things like astrology, ghosts, spontaneous human combus- entertain.” tion, psychic surgery, and ESP. You hear about these fantastic happenings every day on television and in the ERRORS, MEDICINE AND THE LAW. A. Merry and supermarket tabloids. Is any of this true or are they making A. McCall Smith. Cambridge University Press, 2001, it all up? While many people tune just for laughs, plenty of pp. vi + 254, £47.50 Cloth; £17.95 Paper. readers believe their outrageous claims – often because they simply don’t have a clear notion of what science really From the back of the book: “Untoward injuries are is. unacceptably common in medical treatment, at times with “So how do you figure out what constitutes real tragic consequences for patients. The phrases ‘an epidemic science and what is nonsense? Quantum Leaps in the of error’ and ‘the medical toll’ have been coined to describe Wrong Direction carefully deconstructs five examples of this problem of ‘iatrogenic harm’, which it has been pseudoscience – UFOs, out-of-body experiences, astrology, suggested may have contributed to ninety-eight thousand creationism, and ESP – and gives easy recipes to test other deaths per year in the USA. Some of these incidents are the dubious notions so that you can tell what lies in the realm of result of negligence on the part of doctors, but more usually real science and what more properly deserves the tag of they are no more than inevitable concomitants of the pseudoscience. complexity of modern healthcare. This book is fundamentally “Brilliantly illustrated with hilarious cartoons by the about distinguishing the former from the latter. It questions renowned science cartoonist Sidney Harris, Quantum Leaps the understandable, but often inappropriate, tendency to in the Wrong Direction is wonderful fun as well as blame individuals for these events and points out that the illuminating science. This witty, disarmingly delightful book goal of safety is far better served by a sophisticated is for anyone – but especially for those folks who secretly understanding of the difference between negligence and check their horoscope every day.” inevitable error, and by a frank recognition of just why human From a review: “Quantum Leaps is one of a far too error occurs and how things go wrong in any complex small cluster of rational books that respond to ‘voodoo science’. system Wynn and Wiggins carry it out with wit and cunning.” “Although medicine is used as the book’s primary L.M. Lederman, winner of the 1988 Nobel Prize in physics example, the points made apply equally to aviation, many and author of The God Particle. industrial activities, and many other fields of human endeavour. The book advocates a more informed alternative to THE ONE CULTURE? A Conversation About Science. the blaming culture which has increasingly come to J.A. Labinger and H. Collins (Eds.).University of Chicago dominate our response to accidents, whether in the medical Press, 2001, pp. xi + 329, US$17.00/£11.00. field or elsewhere.” From the book cover: “Who has the right to speak CROSSING THE QUALITY CHASM: A New Health System about science? What is the proper role of scientific for the 21st Century. Foreword by the Institute of Medicine. knowledge? How should scientists interact with the rest of Washington, D.C.: National Academy Press, 2001, society in decision making? Because science occupies a pp. xx + 335, US$50.00. central position in the world today, such questions are vitally important. Although there are no simple solutions. The One From the foreword: “This is the second and final Culture? shows the reader exactly what is at stake in the so- report of the Committee on the Quality of Health Care in called Science Wars and provides valuable framework for America. … how to go about seeking the answers we so urgently need.” “The present report addresses quality-related issues more broadly, providing a more strategic direction for 53 redesigning the health care delivery system of the 21st CLIMATE CHANGE. A MULTIDISCIPLINARY APPROACH. century.” W.J. Burroughs. Cambridge University Press, 2001, pp. xv + 298, £52.50/US$85.00 Cloth; £18.95/US$29.95 THE SCIENCE OF COOKING. P. Barham. Berlin: Paper. Springer-Verlag, 2001, pp. 244, US$34.95. From the book cover: “Climate Change: A This book is based on the author’s experiences in Multidisciplinary Approach provides a concise, up-to-date presenting popular lectures in science. presentation of our current knowledge of climate change From the book cover: “A kitchen is no different and its implications for society. The book begins by giving a from most science laboratories and cookery may properly balanced coverage of the physical principles of the global be regarded as an experimental science. Food preparation climate, its behaviour on all timescales, and the evidence and cookery involve many processes which are well for and concequences of past change. It then reviews how described by the physical sciences. Understanding the we measure climate change and the statistical methods for chemistry and physics of cooking should lead to analysing data, before exploring the causes of climate improvements in performance in the kitchen. For those of us change and how we can model this behaviour. The final who wish to know why certain recipes work and perhaps sections discuss predictions of future climate change and more importantly why others fail, appreciating the the economic and political debate surrounding its prevention underlying physical processes will inevitably help in and mitigation.” unravelling the mysteries of the “art” of good cooking.” UNDER THE WEATHER. CLIMATE, ECOSYSTEMS, AND THE CHICAGO GUIDE TO YOUR ACADEMIC CAREER. INFECTIOUS DISEASE. National Research Council. A Portable Mentor for Scholars from Graduate School Washington, D.C.: National Academy Press, 2001, through Tenure. J.A. Goldsmith, J. Komolos and pp. xiv + 146, US$37.95. P. Schine Gold. Chicago, Illinois: University of Chicago From the book cover: “Since the dawn of medical Press, 2001, pp. xx + 310. science, connections between changes in the weather and From the book cover: “Is a career as a professor the the appearance of epidemic diseases have been right choice? With a perpetually tight job market in the recognized. As our understanding of these connections traditional academic fields, the path to an academic career grows, many hope that it will be possible to build models for for many aspiring scholars might become a frustrating one. predicting disease outbreaks based on climate forecasts Where can they turn for good, frank answers to their and ecological observations. However, understanding the questions? Here, three distinguished scholars – with more effects of climate in the context of all of the other forces that than 75 years of combined teaching experience – talk openly drive infectious disease dynamics presents a tremendous about what is good and what is not so good about academia, scientific challenge. Can this understanding be used to help as a place to work and as a way of life. Written as an the field of public health advance from “surveillance and informal conversation among colleagues, the book is packed response” to “prediction and prevention”? And perhaps the with inside information – about finding a mentor, making it question of greatest public concern: Can we predict how through the dissertation, getting a job, obtaining tenure, and global warming will affect the emergence and transmission lots more useful advice.” of infectious disease agents around the world? CLIMATE CHANGE 2001: MITIGATION. Contribution of EVOLUTIONARY GENETICS. From Molecules to Working Group III to the Third Assessment Report of the Morphology. R.S. Singh and C.B. Krimbas (Eds.). Intergovernmental Panel on Climate Change. B. Metz, Cambridge University Press, 2000, pp. xvii + 702, D. Davidson, R. Swart and J. Pan (Eds.). £65.00/US$95.00. Cambridge University Press, 2001, pp. x + 752, From the book cover: “This collection of essays is £90.00/US$130.00 Cloth; £34.95/US$49.95 Paper. produced in honor of Lewontin’s 65th birthday. The volume From the book: “Climate Change 2001: Mitigation is is unique, as it has a comprehensive coverage of modern the most comprehensive and up-to-date scientific, technical evolutionary genetics from molecules to morphology by a and economic assessment of options to mitigate climate group of star authors, including his students and colleagues. change and their costs.” Such a comprehensive treatment of evolutionary genetics has never been attempted before. The volume is set in a CLIMATE CHANGE 2001: THE SCIENTIFIC BASIS. historical perspective, but it has an up-to-date coverage of Contribution of Working Group I to the Third Assessment material in the various fields. The areas covered are the Report of the Intergovernmental Panel on Climate mathematical and the molecular foundations of population Change. J.T. Houghton et al (Eds.). Cambridge University genetics, molecular variation and evolution, selection and Press, 2001, pp. x + 881, £90.00/US$130.00 Cloth; genetic polymorphisms, linkage and breeding system £34.95/US$49.95 Paper. evolution, quantitative genetics and phenotypic evolution, From the book: “Climate Change 2001: The gene flow and population structure, speciation, behavior, Scientific Basis is the most comprehensive and up-to-date and ecology.” scientific assessment of past, present and future climate change.”

NEW EDITIONS, PAPER EDITIONS OR REPRINTS

CORRELATION AND REGRESSION. Applications for 2nd edition. P. Bobko. Thousand Oaks, California: Sage, Industrial Organizational Psychology and Management, 2001, pp. xiv + 287. £47.00 Cloth; £29.00 Paper. 54 55 56

DIFFUSIONS, MARKOV PROCESSES, AND PROBABILITY THEORY, AN ANALYTIC VIEW, MARTINGALES. Volume 2: Itô Calculus, 2nd edition. revised edition. D.W. Stroock. Cambridge University Press, L.C.G. Rogers and D. Williams. Cambridge University 1999, pp. xvi + 536, £20.95/US$29.95 Paper. [Original Press, 2000, pp. xiii + 480, £24.95/US$39.95 Paper. 1993; Short Book Reviews Vol. 14, p. 29] [Original 1987; Short Book Reviews Vol. 8, p. 8] EPIDEMIC MODELLING. D.J. Daley and J. Gani. Cambridge PROBABILITY, RANDOM VARIABLES, AND RANDOM University Press, 2001, pp. xii + 213, £18.95/US$27.95 SIGNAL PRINCIPLES, 4th edition. P.Z. Peebles Jr. Paper. New York: McGraw-Hill, 2001, pp. xviii + 462. [Original 1999, Short Book Reviews Vol. 19, p. 29] MATHEMATICAL STATISTICS WITH APPLICATIONS, THE MATHEMATICS OF COMPUTERIZED 6th edition. D.D. Wackerly, W. Mendenhall III and TOMOGRAPHY. F. Natterer. Philadelphia: Society for R.L. Scheaffer. Australia: Duxbury, 2002, pp. xvi + 853. Industrial and Applied Mathematics, 2001, pp. xvii + 222. [Original, 1986]

GOVERNMENT PUBLICATIONS

NETHERLANDS OFFICIAL STATISTICS. Volume 15, QUARTERLY NATIONAL ACCOUNTS MANUAL. Concepts, Winter 2000. Voorburg: Statistics Netherlands, 2000, Data Sources, and Compilation. A.M. Bloem, pp. 37, DFl20.00/€9.08 per copy/DFl42.00/€19.06. Annual R.J. Dippelsman and N.Ø. Mæhle. Washington, D.C.: subscription. International Monetary Fund, 2001, pp. xii + 210, US$40.00.

UNITED NATIONS STATISTICAL OFFICE PUBLICATIONS RECENTLY ISSUED

COMPENDIUM OF HUMAN SETTLEMENTS STATISTICS Data available as of 1 January 2001. 2001. Sixth issue. ST/ESA/STAT/SER.N/6, 2001, ST/ESA/STAT/SER.A/216, 2001, pp. 19. U.N. Sales No. E.01.XVII.5, pp. v + 246 POPULATION AND VITAL STATISTICS REPORT. Statistical 1998 DEMOGRAPHIC YEARBOOK. Fiftieth issue. Papers series A.Vol. LIII, No. 2. ST/ESA/STAT/SER.R/29, 2000, Data available as of 1 April 2001, U.N.Sales No. E/F.00.XIII.1, pp. viii + 610. ST/ESA/STAT/SER.A/217, 2001, pp. 19.

HANDBOOK ON CENSUS MANAGEMENT FOR RESULTS OF THE EIGHT UNITED NATIONS INQUIRY POPULATION AND HOUSING CENSUSES. Studies in AMONG GOVERNMENTS ON POPULATION AND Methods series F No. 83/Rev. 1 DEVELOPMENT. ST/ESA/SER.R/155, 2001, ST/ESA/STAT/SER.F/83/Rev. 1, 2000, U.N.Sales No. E.01.XIII.2. pp. xii + 198. U.N.Sales No. E.00.XVII 15 Rev. 1, pp. ix + 261. TOURISM, SATELLITE ACCOUNT: Recommended 1998 INDUSTRIAL COMMODITY STATISTICS YEARBOOK. Methodological Framework. Commission of the European Production Statistics (1989–1998). Communities, Organisation for Economic Co-operation and ST/ESA/STAT/SER.P/38, 2000, Development, United Nations and World Tourism U.N.Sales No. E/E.01 XVII 3, pp. xvii + 889. Organisation. Luxembourg/Madrid/New York/Paris: C.E.C., O.E.C.D., W.T.O., U.N., O.E.C.D. POPULATION AND VITAL STATISTICS REPORT. Statistical Code 78 2000 01 IPI, U.N.Sales No. E. 01 XVII 9, Papers series A. Vol. LIII, No. 1. ST/ESA/STAT/SER.E/80. 2001, pp. xii + 138.

NEW JOURNALS

STATISTICAL MODELLING. An International Journal. £75.00 (Individual) Annual subscription (4 issues); £60.00 Volume 1, Number 1, April 2001. J. Hinde, E. Lesaffre and Single issue. B. Marx (Eds.). London: Arnold Journals, 2001, pp. 80,

COLLECTED PAPERS, TABLES AND PROCEEDINGS

ADVANCED MEAN FIELD METHODS. Theory and Practice. CONTROL OF NONLINEAR DISTRIBUTED PARAMETER M. Opper and D. Saad (Eds.). Cambridge, Massachusetts: SYSTEMS. G. Chen, I. Lasiecka and J. Zhou (Eds.). The MIT Press, 2001, pp. xiii + 273, £27.50. New York: Dekker, 2001, pp. xii + 357, US$150.00. APPLIED STATISTICS IN THE PHARMACEUTICAL INDUSTRY. With Case Studies Using S-PLUS. S.P. Miljard FLUCTUATIONS AND SCALING IN BIOLOGY. T. Vicsek and A. Krause (Eds.). New York: Springer-Verlag, 2001, (Ed.). Oxford University Press, 2001, pp. ix + 246, £50.00 pp. xviii + 513, US$79.95/DM182.00. Cloth; £24.95 Paper. 57

HUNGARIAN PROBLEM BOOK III. Based on the Eotvos PUBLIC HEALTH SYSTEMS AND EMERGING Competition 1929–1943. Compiled by G. Hajos, INFECTIONS. Assessing the Public and Private Sectors. G. Neukomm and J. Suranyi. A. Liu (Ed.). Workshop Summary. J.R. Davis and J. Lederberg (Eds.). Washington, D.C.: The Mathematical Society of America, Washington, D.C.: National Academy Press, 2000, 2001, pp. xviii + 142, US$27.95. pp. xvi + 110, US$23.00. INDEPENDENT COMPONENT ANALYSIS: PRINCIPLES RANDOM WALKS AND DISCRETE POTENTIAL THEORY. AND PRACTICE. S. Roberts and R. Everson (Eds.). Cortona 1997. M. Picardello and W. Woess (Eds.). Cambridge University Press, 2001, pp. xii + 338, Cambridge University Press, 1999, pp. viii + 361, £45.00/US$69.95. £47.50/US$69.95. MATHEMATICAL FINANCE. Workshop of the Mathematical TOXICOLOGICAL EFFECTS OF METHYLMERCURY. Finance Research Project, Konstanz, Germany, October 5- National Research Council. Washington, D.C.: 7, 2000. M. Kohlmann and S. Tang (Eds.). Basel, National Academy Press, 2000, pp. xvii + 344, US$54.00. Switzerland: Birkhäuser, 2001, pp. 374, SFr148.60/DM196.00/ÖS431.00. USA AND INTERNATIONAL OLYMPIADS 2000. MODERN SAMPLING THEORY. Mathematics and T. Andreescu and Z. Feng (Eds.). Washington, D.C.: The Applications. J.J. Benedetto and P.J.S.G. Ferreira (Eds.). Mathematical Association of America, 2001, pp. xiv + 83, Boston: Birkhäuser, 2001, pp. xvi + 417, US$16.50. SFr168.00/DM220.00/ÖS1606.00.

BOOKS RECEIVED

A FIRST COURSE IN PROBABILITY. T.K. Chandra and EXPLORING THE BIOLOGICAL CONTRIBUTIONS TO D. Chatterjee. New Delhi: Narosa, 2001, pp. xii + 467, HUMAN HEALTH. DOES SEX MATTER? Institute of £45.00. Medicine. Washington, D.C.: National Academy Press, 2001, pp. xv + 206, US$45.00. A FIRST COURSE ON PARAMETRIC INFERENCE. B.K. Kale. New Delhi: Narosa, 1999, pp. x + 268, £29.50. FUNDAMENTALS OF APPROXIMATION THEORY. H.N. Mhaskar and D.V. Pai. Boca Raton, Florida: CRC, ASTRONOMY AND ASTROPHYSICS IN THE NEW New Delhi: Narosa, 2000, pp. xv + 541. MILLENNIUM. National Research Council. GEOMETRIC MODELING WITH SPLINES: AN Washington, D.C.: National Academy Press, 2001, INTRODUCTION. E. Cohen, R.F. Riesenfeld and G. Elber. pp. xxiv + 246, US$34.95. Natick, Massachusetts: Peters, 2001, pp. xxii + 616, BASIC STATISTICS AND PHARMACEUTICAL US$59.99/£40.00. STATISTICAL APPLICATIONS. J.E. de Muth. New York: HUMAN PERCEPTION OF OBJECTS. Early Visual Dekker, 1999, pp. xii + 596, US$79.75. Processing of Spatial Form Defined by Luminance, Color, Texture, Motion, and Binocular Disparity. D. Regan. BIOMETRIC MODELLING WITH SPLINES: An Introduction. Sunderland, Massachusetts, 2000, pp. xxix + 577, £35.99. E. Cohen, R.F. Riesenfeld and G. Elber. Natick, Massachusetts: Peters, 2001, pp. xxii + 616, INTRODUCTION TO GAUGE INTEGRALS. C. Swartz. US$59.99/£40.00. Singapore: World Scientific, 2001, pp. x + 157, US$38.00. INTRODUCTION TO NUMERICAL ANALYSIS. A. Neumaier. CONTROL SYSTEMS THEORY WITH ENGINEERING Cambridge University Press, 2001, pp. viii + 356, APPLICATIONS. S.E. Lyshevski. Boston: Birkhäuser, £65.00/US$95.00 Cloth; £23.95/US$95.00 Paper. 2001, pp. xi + 416 + CD. SFr148.00/DM196.00/ ÖS1431.00. MATHEMATICAL OLYMPIAD CHALLENGES. T. Andreescu and R. Gelca. Boston: Birkhäuser, 2000, pp. xv + 260, COPPER IN DRINKING WATER. National Research Council. SFR118.00/DM138.00/ÖS1008.00. Washington, D.C.: National Academy Press, 2000, pp. xii + 147, US$35.00. MIND ON STATISTICS. J.M. Utts and R.F. Heckard. Australia: Duxbury Thompson Learning, 2002, COUNTING ON FRAMEWORKS. Mathematics to Aid the pp. xxi + 568 + CD. Design of Rigid Structures. J.E. Graver. Washington, D.C.: MULTIFRACTALS. Theory and Applications. D. Harte. The Mathematical Association of America, 2001, Boca Raton, Florida: Chapman and Hall/CRC, 2001, pp. xii + 180, US$28.95. pp. xiii + 248, US$79.95/£44.99. DIFFERENTIAL GAMES IN ECONOMICS AND NONLINEAR DYNAMICAL SYSTEMS. Feedforward Neural MANAGEMENT SCIENCE. E.J. Dockner, S. Jørgensen, Network Perspectives. I.W. Sandberg, J.T. Lo, N. van Long and G. Sorger. Cambridge University Press, C.L. Fancourt, J.C. Principe, S. Katagiri and S. Haykin. 2000, pp. xii + 382, £55.00/US$90.00 Cloth; New York: Wiley, 2001, pp. x + 298, £57.50. £19.95/US$31.95 Paper. NON-PHOTOREALISTIC RENDERING. B. Gooch and A. Gooch. Natick, Massachusetts: Peters, 2001, ECONOMIC FOUNDATIONS. R.C. Mittelhammer, pp. x + 243, US$39.00/£27.00. G.G. Judge and D.J. Miller. Cambridge University Press, 2000, pp. xxviii + 756 + disk. ON SOCIOLOGY. NUMBERS, NARRATIVES, AND THE INTEGRATION OF RESEARCH AND THEORY. ENHANCING DATA SYSTEMS TO IMPROVE THE J.H. Goldthorpe. Oxford University Press, 2000, QUALITY OF CANCER CARE. National Research Council. pp. vi + 337, £16.99 Paper. Washington, D.C.: National Academy Press, 2000, pp. xii + 163, US$30.00. 58 59

PRACTICAL GEOSTATISTICS 2000. I. Clark and SOLVE THIS. Math Activities for Students and Clubs. W.V. Harper. Columbus, Ohio: Ecosse North America, J.S. Tanton. Washington, D.C.: The Mathematical 2000, pp. 342 + CD. Association of America, 2001, pp. xiii + 218, US$29.95. PROBABILITY APPLICATIONS IN MECHANICAL DESIGN. STATISTICAL DATA ANALYSIS USING YOUR PERSONAL F.E. Fisher and J.R. Fisher. New York: Dekker, 2000, COMPUTER. I.H. Bernstein and N.A. Rowe. pp. xi + 276, US$135.00. Thousand Oaks, California: Sage, 2001, pp. ix + 446, £40.00 Cloth; £18.99 Paper. QUANTITATIVE APPROXIMATIONS. C. Anastassiou. Boca Raton, Florida: Chapman and Hall/CRC, 2001, STATISTICAL THINKING. IMPROVING BUSINESS pp. xiv + 607. PERFORMANCE. R.W. Hoerl and R.D. Snee. Australia: Duxbury, 2002, pp. xvii + 526. RANDOM WALKS ON INFINITE GRAPHS AND GROUPS. W. Woess. Cambridge University Press, 2000, pp. xi + 334, THE GOLDEN SECTION. H. Walser. Washington, D.C. The £42.50/US$64.95. Mathematical Association of America, 2001, pp. xvi + 142, US$26.95. REGULAR SEQUENCES AND RESULTANTS. G. Scheja and U. Storch. Natrick, Massachusetts: Peters, 2001, THE INTERNET’S COMING OF AGE. National Research pp. viii + 142, US$30.00/£21.00. Council. Washington, D.C.: National Academy Press, 2001, pp. xvii + 236, US$34.95. RISK ANALYSIS AND UNCERTAINTY IN FLOOD DAMAGE REDUCTION STUDIES. National Research Council. TREES: THEIR NATURAL HISTORY. P. Thomas. Washington, D.C.: National Academy Press, 2000, Cambridge University Press, 2001, pp. ix + 286. pp. xiv + 202, US$40.00. £50.00/US$69.95 Cloth; £16.95/US$24.95 Paper. SAS PROGRAMMING FOR RESEARCHERS AND SOCIAL TWO JEWS ON A TRAIN. Stories from the Old Country and SCIENTISTS. 2nd edition. P.E. Spector. Thousand Oaks, the New. A. Biro. University of Chicago Press, 2001, California: Sage, 2001, pp. xv + 231, £47.00 Cloth; £25.00 pp. xiv + 128, US$17.00/£11.00. Paper. WAVELET METHODS – ELLIPTIC BOUNDARY VALUE SMALL-AREA INCOME AND POVERTY ESTIMATES. PROBLEMS AND CONTROL PROBLEMS. A. Kunoth. Priorities for 2000 and Beyond. National Research Council. Stuttgart: Teubner, 2001, pp. 141, Washington, D.C.: National Academy Press, 2000, DM78.00/ÖS569.00/SFr68.50. pp. xii + 206, US$32.00.

INDEX VOLUME 21, 2001

Andel, J. Mathematics of Chance. (F.H Berkshire) p. 42. Demuth, M. and Casteren, J.A. Stochastic Spectral Theory Asmussen, S. Ruin Probabilities. (H. Schmidli) p. 9. for Self-Adjoint Feller Operators: A Functional Integration Aste, T. and Weaire, D. The Pursuit of Perfect Packing. Approach. (N.H. Bingham) p. 50. (F.H. Berkshire) p. 1. Devroye, L. and Lugosi, G. Combinatorial Methods in Density Atkinson, A. and Riani, M. Robust Diagnostic Regression Estimation. (N.D.C. Veraverbeke) p. 23. Analysis. (N.R. Draper) p. 4. Dodge, Y. and Jureckova, J. Adaptive Regression. Balding, D.J., Bishop, M. and Cannings, C. (Eds.). Handbook (V.N. Vapnik) p. 4. of Statistical Genetics. (D. Stephens) p. 44. Does, R.J.M.M., Roes, K.C.B. and Trip, A. Statistical Process Barndorff-Nielsen, O.E., Cox, D.R. and Klüppelberg, C. (Eds.). Control in Industry: Implementation and Assurance of SPC. Complex Stochastic Systems. (D.J. Hand) p. 31. (C.A. Fung) p. 3. Barndorff-Nielsen, O.E., Mikosch, T. and Resnick, S.I. Doucet, A., de Fretas, N. and Gordon, N. (Eds.). Sequential (Eds.). Lévy Processes: Theory and Applications. Monte Carlo Methods in Practice. (D.L. McLeish) p. 49. (P.A.L. Embrechts) p. 49. Durbin, J. and Koopman, S.J. Time Series Analysis by State Bedford, T. and Cooke, R. Probabilistic Risk Analysis: Space Methods. (P. Whittle) p. 48. Foundations and Methods. (P.A.L. Embrechts) p. 47. Everitt, B.S. and Dunn, G. Applied Multivariate Data Analysis, Campbell, J. Rutherford: Scientist Supreme. (D.A. Ramsay) 2nd edition. (B.J.T. Morgan) p. 45. p. 21. Everitt, B.S., Landau, S. and Leese, M. Cluster Analysis. Carter, M.W. and Price, C.C. Operations Research: A 4th edition (W.J. Krzanowski) p. 45. Practical Introduction. (S. Powell) p. 32. Fitzgerald, W.J., Smith, R.L, Walden, A.T. and Young, P.C. Charalambides, C.A., Koutras, M.V. and Balakrishnan, N. (Eds.). Non-linear and Nonstationary Signal Processing. (Eds.). Probability and Statistical Models with Applications. (T. Subba Rao) p. 31. (A. Parsian) p. 29. Fouque, J.P., Papanicolaou, G. and Sircar, K.R., Derivatives Chatfield, C. Time Series Forecasting. (R. Coleman) p. 48. in Financial Markets with Stochastic Volatility. (P.A.L. Embrechts) p. 9. Chen, J. and Gupta, A.K. Parametric Statistical Change Point Franses, P.H. and van Dijk, D. Nonlinear Time Series Models Analysis. (V. Dragalin) p. 4. in Empirical Finance. (R. Coleman) p. 9. Clewer, A.G. and Scarisbrick, D.H. Practical Statistics and French, S. and Rios Insua, D. Statistical Decision Theory. Experimental Design for Plant and Crop Science. (J.V. Zidek) p. 28. (P. Prescott) p. 25. Gail, M.B. and Benichou, J. Encyclopedia of Epidemiological Clote, P. and Backhofen, R. Computational Molecular Methods. (B.L. De Stavola) p. 42 Biology: An Introduction. (A.W. Kemp) p. 7. Gilchrist, W.G. Statistical Modelling with Quantile Functions. Congdon, P. Bayesian Statistical Modelling. (D. Stephens) (A. Parsian) p. 3. p. 46. Hosmer, D.W. and Lemeshow, S. Applied Logistic Cox, T.F. and Cox, M.A.A. Multidimensional Scaling, 2nd Regression, 2nd edition. (C.M. O’Brien) p. 27. edition. (D.J. Hand) p. 27. Hougaard, P. Analysis of Multivariate Survival Data. David, H.A. and Edwards, A.W.F. Annotated Readings in the (J.M. Juritz) p. 6. History of Statistics. (N.R. Draper) p. 41. 60

Jacod, J. and Protter, P. Probability Essentials. Pistone, G., Riccomagno, E. and Wynn, H. Algebraic (P.A.L. Embrechts) p. 23. Statistics: Computational Commutative Algebra in Jagerman, D.L. Difference Equations with Applications to Statistics. (D.F. Andrews) p. 28. Queues. (A.W. Kemp) p. 31. Politis, D.N., Romano, J.P. and Wolf, M. Subsampling. Janacek, G. Practical Time Series. (T. Subba Rao) p. 25. (V.V. Fedorov) p. 4. Jouini, E., Cvitanić, J. and Musiela, M. (Eds.). Handbooks in Press, S.J. and Tanur, J.M. The Subjectivity of Scientists and Mathematical Finance. Option Pricing, Interest Rates and the Bayesian Approach. (D.J. Hand) p. 42. Risk Management. (D.J. Hand) p. 47. Quinn, B.G. and Hannan, E.J. The Estimation and Tracking of Kageyama, S. and Calinski, T. Block Designs. Volume 1: A Frequency. (A.T. Walden) p. 48. Randomization Approach. (P.W.M. John) p. 44. Rao, M.M. Stochastic Processes. Inference Theory. Kaplan, D. Structural Equation Modeling: Foundations and (N.D.C. Veraverbeke) p. 7. Extensions. (C.M. O’Brien) p. 28. Rao, P.S.R.S. Sampling Methodologies with Applications. (K.- Karian, Z.A. and Dudewicz, E.J. Fitting Statistical W. Tsui) p. 24. Distributions. (N.R. Draper) p. 24. Rigdon, S.E. and Basu, A.P. Statistical Methods for the Kemp, K.W. The Efficient Use of Quality Control Data. Reliability of Repairable Systems. (R.G. Jarrett) p. 30. (C.A. Fung) p. 49. Rohatgi, V.K. and Saleh, A.K.M.E. An Introduction to Korn, R. and Korn, E. Option Pricing and Portfolio Probability and Statistics, 2nd edition (S. Starkings) p. 43. Optimization, Modern Methods of Financial Mathematics. Rutherford, A. Introducing ANOVA and ANCOVA – A GLM (P. Vanini) p. 32. Approach. (C.M. O’Brien) p. 26. Kuo, W., Prasad, V.R., Tillman, F.A. and Hwang, C.-L. Ryan, T.P. Statistical Methods for Quality Improvement. Optimal Reliability Design. (M. Crowder) p. 30. (C.A. Fung) p. 2. Lawson, A.B. Statistical Methods in Spatial Epidemiology. Salsburg, D. The Lady Tasting Tea. How Statistics (D. Stephens) p. 43. Revolutionized Science in The Twentieth Century. Le Cam, L. and Lo Yang, G. Asymptotics in Statistics. Some (G.A. Barnard) p. 41. Basic Concepts, 2nd edition. (N.D.C. Veraverbeke) p. 7. Severini, T.A. Likelihood Methods in Statistics. (P.J. Laycock) Lele, R.S. and Richtsmeier, J.T. An Invariant Approach to p. 46. Statistical Analysis of Shapes. (C.G. Small) p. 23. Shorack, G.R. Probability for Statisticians. (D.L. McLeish) p. 2. Limnios, N. and Oprisan, G. Semi-Markov Processes and Sprent, P. and Smeeton, N.C. Applied Nonparametric Reliability. (R. Pyke) p. 50. Statistical Methods, 3rd edition. (J.M. Juritz) p. 22. Manly, B.F.J. Statistics for Environmental Science and Sprott, D.A. Statistical Inference in Science. (E.L. Lehmann) Management. (J.M. Juritz) p. 22. p. 7. Mari, D.D. and Kotz, S. Correlation and Dependence. Steele, J. Stochastic Calculus and Financial Applications. (N.R. Draper) p. 46. (D.L. McLeish) p. 9. Matthews, J.N.S. An Introduction to Randomized Controlled Sutton, A.J., Abrams, K.R., Jones, D.R., Sheldon, T.A. and Clinical Trials. (C.M. O’Brien) p. 43. Song, F. Methods for Meta-Analysis in Medical Research. McCulloch, C.E. and Searle, S.R. Generalized, Linear, and (V.T. Farewell) p. 6. Mixed Models. (P. Prescott) p. 26. Taniguchi, M. and Kakizawa, Y. Asymptotic Theory of Meyer, M. Continuous Stochastic Calculus with Applications Statistical Inference for Time Series. (N.D.C. Veraverbeke) to Finance. (D.L. McLeish) p. 47. p. 8. Miles, J. and Shevlin, M. Applying Regression and Therneau, T.M. and Gambsch, P.M. Modelling Survival Data: Correlation: A Guide for Students and Researchers. Extending The Cox Model. (J.M. Juritz) p. 5. (N.R. Draper) p. 27. Thorrison, H. Coupling, Stationarity, and Regeneration. Miller, R.E. Optimization: Foundations and Applications. (W.S. Kendall) p. 8. (S. Powell) p. 10. Valliant, R., Dorfman, A.H. and Royall, R.M. Finite Population Mollin, R.A. An Introduction to Cryptography. (N.R. Draper) Sampling and Inference. A Predictive Approach. p. 21. (P. Prescott) p. 45. Morton, S.C. and Rolph, J.E. Public Policy and Statistics: van Lieshout, M.N.M. Markov Point Processes and their Case Studies from Rand. (R.W. Oldford) p. 1. Applications. (R. Pyke) p. 8. Webb, A. Statistical Pattern Recognition. (M.F. Ramalhoto) Mukhopadhyay, N. Probability and Statistical Inference. p. 5. (C.D. Kemp) p. 24. Webster, R. and Oliver, M.A. Geostatistics for Environmental Nash, J.C. and Nash M.M. Practical Forecasting for Scientists. (P.J. Laycock) p. 44. Managers, (S. Starkings) p. 48. Whittle, P. Probability via Expectation, 4th edition. Ohser, J. and Mücklich, F. Statistical Analysis of (P.A.L. Embrechts) p. 2. Microstructures in Material Science. (J.D. Picka) p. 30. Wilcox, R.R. Fundamentals of Modern Statistical Methods. Peña, D., Tiao, G.C. and Tsay, R.S. (Eds.). A Course in Time (A. Parsian) p. 22. Series Analysis. (M. Crowder) p. 25. Yoshihara, K. Weakly Dependent Stochastic Sequences and Percival, D.B. and Walden, A.T. Wavelet Methods for Time their Applications. Volume XI: Censorship under Weak Series Analysis. (T. Subba Rao) p. 28. Dependence. (N.D.C. Veraverbeke) p. 32.