Excellent! The focus is on statistics in medicine, but the book zigzags through recent issues (ethics and politics of clinical trials, lawyer's abuse of statistical evidence, vaccine scares), sometimes sophisticated analysis of particular data, combined with explanation and history of basic concepts, with half-page biographies of historical and modern statisticians going far beyond the usual suspects. Has the lively style of The Economist, addressing a mentally alert adult reader rather than a casual reader or bored student. In particular, readers who have taken one course in statistics will get a view of "the big picture", and this is the best single book for that purpose.
Senn provides a broad introduction to key statistical ideas relevant to medicine and epidemiology. He switches effortlessly between basic concepts such as hypothesis testing, standard error and conditional probabilities to deep philosophy such as Bayesian versus frequentist schools, irrationality of induction and the use of meta-analysis. His style is highly readable mixing technical content with historical anecdotes and startling digressions.
The ideal reader is someone who has a decent background in statistics, such as that gained in a university-level statistics course.
The key strengths of the book include: (1) clear, lucid explanation of many concepts including modern ones not typically covered in first courses -- for instance, the contrast between Bayesian and frequentist approaches is done much better here than in most other books; (2) about 1/2-1/3 of the examples come from the real world (the remainder split between coin tossing experiments and hypotheticals), which is a high proportion among this sort of books; (3) good discussion of intuition and reasoning as opposed to just formulae.
The weaknesses (quibbles) include: (1) Senn's penchant for puns and word play provides humor but can get in the way of understanding the material; (2) his frequent digressions leave a host of loose ends and dead ends, which can frustrate some readers but for others, this strategy reveals exciting avenues for further exploration.
This is really a great book for someone who knows some statistics and really want to dig much deeper into the intuition and philosophy behind this field.
The title of the book and the first page of the first chapter gave me a thrill, and I thanked my friend who introduced me this book. There are plenty of examples and anecdotes that cannot be found in usual statistical books. The author tries to cover many topics in statistics, ranging from the concept of probability to experimental designs. However, I do not know the target reader of the book. If it is for the novice statisticians, writings will be too complicated; if it is for the educated statisticians, explanation is too wordy. For example, the author tries to verbally explain marginal, conditional and joint probability using a 2 by 2 contiongency table spending a whole 1 page. Novice statisticians will find the explanation unnecessarily complicated; the educated will find it unnecessarily lengthy and redundant. Some, if not most, examples are quite complicated and cannot be grasped just by reading the book unless the reader is already well familiar with the concept of the examples. If the reader is already familiar with the examples, he/she will find them trivial; if not, will later find out that the reader has spent too much time unnecessarily to understand them.
I have just completed 2 semesters of graduate level biostats and I find this book really dense and excessively wordy. I think this is a cultural issue being that he is from England. When I bought the book I was hoping for a Malcolm Gladwell type of breakdown of medical statistics. Unfortunately, the author doesn't have Malcolm's ability to tell a really good story using high level theories.
I bought this book first to read on a long trip and second to reinforce the knowledge I have worked so hard to understand. Since I am going on a long trip I gambled on this book and lost. Now I am stuck with one less book.