Intuitive Biostatistics is a beautiful book that has much to teach experimental biologists of all stripes. Unlike other statistics texts I have seen, it includes extensive and carefully crafted discussions of the perils of multiple comparisons, warnings about common and avoidable mistakes in data analysis, a review of the assumptions that apply to various tests, an emphasis on confidence intervals rather than P values, explanations as to why the concept of statistical significance is rarely needed in scientific work, and a clear explanation of nonlinear regression (commonly used in labs; rarely explained in statistics books).
In fact, I am so pleased with Intuitive Biostatistics that I decided to make it the reference of choice for my postdoctoral associates and graduate students, all of whom depend on statistics, and most of whom need a closer awareness of precisely why. Motulsky has written thoughtfully, with compelling logic and wit. He teaches by example what one may expect of statistical methods and perhaps just as importantly, what one may not expect of them. He is to be congratulated for this work, which will surely be valuable and perhaps even transformative for many of the scientists who read it.
—Bruce Beutler, 2011 Nobel Laureate, Physiology or Medicine
Director, Center for the Genetics of Host Defense, UT Southwestern Medical Center
After struggling with books that weren’t right for my class, I was delighted to find Intuitive Biostatistics. It is the best starting point for undergraduate students seeking to learn the fundamental principles of statistics because of its unique presentation of the important concepts behind statistics. Lots of books give you the “recipe” approach, but only Intuitive Biostatistics explains what it all means. It meticulously goes through common mistakes and shows how to correctly choose, perform, and interpret the proper statistical test. It is accessible to new learners without being condescending.
—Beth Dawson, The University of Texas at Austin
This textbook emphasizes the thinking needed to interpret statistical analysis in published research over knowledge of the mathematical underpinnings. The basics of choosing tests and doing simpler analyses are covered very clearly and simply. The language is easy to understand yet accurate. It brings in the higher level of intuitive understanding that we hope students will have at the end of an honors undergraduate or MSc. program, skipping over the mathematical details that are now handled by software, anyway. It is the prefect approach and level for undergraduates beginning research.
—Janet E. Kübler, Biology Dept., California State University at Northridge
I’ve read several statistics books, but found that some concepts I was interested in were not mentioned and other concepts were hard to understand. You can ignore the “bio” in Intuitive Biostatistics, as it is the best applied statistics books I have come across, period. Its clear, straightforward explanations have allowed me to better understand research papers and select appropriate statistical tests. Highly recommended.
—Ariel H. Collis, Economist, Georgetown Economic Services
Intutitive Biostatistics provides students with a friendly, gentle introduction to a much maligned subject. Motulsky has done an impressive job of bringing life to biostatistics and allowing the next generation of students to welcome the growing quanitative approach to biology. This work even introduces some of the more advanced statistical techniques such as linear regression, nonlinear regression and ANOVA. Motulsky’s work also provides a nice review of basic topics for more advanced students who are in need of brushing up before tackling more advanced courses.
—Philip Hejduk, University of Texas at Arlington
"I wish everyone involved in medical and pharmaceutical research would read Intuitive Biostatistics: A Nonmathematical Guide to Statistical Thinking. That would likely save us all a great deal of money, and cut way back on nonsense promulgated as new science. . . .
The problem is that these course seldom teach any understanding of the assumptions behind the statistical models the students learn to generate. There is often no mention of Bayesian analysis and its assumptions and conclusions. Mostly the students learn to look for "significance" and generally find it. ...The problem is that the statistical models used in determining the "significance" level of a result involve assumptions that are often not met in the experimental design and this is because the experimenter hasn't been taught much about statistical inference and its assumptions. This can be costly. . . .
Intuitive Biostatistics takes a different approach. There is very little about cookbook techniques in this book; instead it concentrates on the underlying assumptions, and just why some inferences are valid and some are not. It's not easy going, but statistical inference isn't an easy subject. Alas, now that computers have made it easy to do the cookbook work to compile all kinds of statistical number, many think that the subject is easy, and computers can do all the work. Intuitive Biostatistics will disabuse them of that error. The book uses many examples from pharmacology research. Recommended.
“This splendid book meets a major need in public health, medicine, and biomedical research training -- a user-friendly biostatistics text for non-mathematicians that clearly explains how to make sense of statistical results, and how to avoid being confused by statistical nonsense. You may enjoy statistics for the first time!”
--Gilbert S.Omenn, Professor of Medicine, Genetics, Public Health, and Computational Medicine & Bioinformatics, University of Michigan
"My copy of Motulsky's book is not at hand. I recommended and lent it to a friend who seems reluctant to return it. The book is not for dummies, but rather for intelligent, curious people who want to know what statistics is all about without getting bogged down in mathematical details. And my friend, who is obviously fascinated by the book, is both intelligent and curious. The second edition is a substantial improvement. I am particularly impressed by the chapters on multiple comparisons. This is increasingly important for such molecular trickery as gene expression chips, which produce a very large number of possible comparisons. Individual comparisons and even a Bonferroni correction are often inadequate. Motulsky deals with a newer method, false discovery rate (FDR), in a clear, understandable way. I'll be recommending the new edition with even greater enthusiasm."
--James F. Crow, Professor Emeritus of Genetics, University of Wisconsin
Its main strengths are the explanation of the very basic concepts such as confidence intervals, P values, standard deviation, survival curves, relative risk, (a little of) odds ratio, clinical/screening tests, etc. To be honest, I only began to really understand and differentiate all these concepts after reading this book. It is very clear, easy to understand, and very repetitive (in a good way). Now I feel so more confident in interpreting research papers (with all those overwhelming P values) and stats outputs.
Motulsky spends a great deal of chapters explaining the basics I mentioned earlier. Therefore, if what you need is something on statistic tests such as ANOVA, MANOVA, regression, etc., this book wouldn't be very helpful for you. Although it does try to explain how to interpret the output, the F value and alike, it is very, very superficial. In this case you would better off getting a more advanced statistics book.
But it definitely fitted my needs as a graduate student, and it does deserve five stars to me.
--Laura Gougeon, on amazon.ca
This is a first-rate book on biostatistics. Strongly recommended for physicians and others who read and must make sense of the biomedical literature. It should also be required reading for scientists who are publishing biomedical research, preferably early in their careers when they are still impressionable.
- Clearly shows the strengths and limitations of statistics in interpreting research results.
- Does a great job on the ever-present problem of multiple comparisons and retrospective data torture.
- Other texts are cookbooks which show you how to perform various statistical tests, but ultimately it is far more important to understand the rationale behind statistical methodology and to see what statistical tests can--and can't--tell us.
- Enforces the notion that statistical tests have assumptions and describes when these assumptions can be violated and when such violations render the tests worthless.
- Debunks the notion that "statistical significance" is an absolute concept and shows how p<0.05 is a convention, and an arbitrary one at that.
- Nice section on the weakness of correlation coefficients for data which violate the assumptions of that test.
- Strong on applying Bayesian thinking to marginal p values.
- Points out the important distinction between clinical and statistical significance.
All in all, a great book.
--Benjamin Ephraim (posted on amazon.com)
I wanted to write and let you know that I am entranced by the book. Statistics is a topic that is often difficult for many scientists to fully appreciate. Your writing style and explanation makes the concepts accessible. Thank you for writing it.
--Tim Bushnell, Director of Flow Cytometry, Univ. Rochester Med. Center
"A great (p<0.0001) intro. This is a must for every medical student. For those engaging in research there is sufficient detail to be able to apply the correct statistical technique and to interpret the results. For the rest it gives a very clear introduction to biostatistics and particularly aids understanding of research articles without which a doctor will quickly lose touch with evidence based medicine.
For the scientist and science student I would recommend this as an adjunct to any statistical course and as a reference to have on hand when reading those articles. I regularly picked up the first edition to check I was understanding something correctly - this edition is an order of magnitude better. "
-- Kiwiski (posted on amazon.com)
"Jargon free statistics. This book fills a gap in the market of books for statistics. The book focuses on the application of statistical procedures, rather than the mathematics of the various tests. The book is written in a non-complicated, conversational way and is filled with interesting examples. The author is able to make a potentially complicated topic, easy to understand and logical. This book is a must for any postgraduate student/professional scientist who uses statistics."
--M.I. Lambert (posted on amazon.com)
"The book's title suggests that he can make biostatistics intuitive for non-statisticians (e.g. physicians, clinicians and nurses). After reading through it he has made a believer out of me! He introduces concepts through examples and touches on most of the important statistical methods that are used in the medical literature. ... My usual concern with such books is that concepts are oversimplified and the presentation is too cook-bookish. Amazingly that is not the case here. Motulsky carefully explains concepts such as confidence intervals, p-values, multiple comparison issues, Bayesian thinking and Bayesian controversy in a way that should be understandable to his intended audience."
--Michael R. Chernick, PhD (review posted on amazon.com)
"I highly recommend this book for those needing a non-mathematical, explanatory introduction to biostatistics. It is well-written and provides wonderful clinical examples and biostatistical content...An excellent resource book for medical students and housestaff who are struggling along with the concepts; and for those of you who were wondering, it was surprisingly easy to read.
--Joseph Chu, book review of Intuitive Biostatistics, Teaching and Learning Medicine, 9:243, 1997
"Let me congratulate you for this unique jewel. I am a biologist, and always found statistics to be a bit confusing. Thanks to your book, statistics now makes sense. Top marks for an excellent work. I will recommend it highly to my students and colleagues."
--Gabriel Dorado, Dept. of Biochemistry and Molecular Biology, University of Cordoba, Spain