Statistical Models

At this point I have read most, but not all of this book. This book is well organized and covers many foundational topics in modern mathematical statistics. This is not a light read, but is worth the work to gain deeper understanding of statistical analysis methods. This book targets people with graduate level statistics and mathematics knowledge. If this level of book is what you are looking for, I can recommend this book along with Efron and Hastie.

This is a wonderful book to read, after you know what you are doing. I simply can't imagine trying to teach from this book, at least not in the United States.

The book is well worth reading, especially for those of us who are not well versed in matematics and matematical formalism. Some background in basic frequentist and/or Bayesian statistics is needed. Otherwise the book is easy to read and the real life application of the examples is easy to apply.

Anthony Davison is one of the rare applied statisticians with in-depth knowledge in statistical theory. He is an excellent researcher and lecturer. Anthony is the editor of Biometrika, one of the most prestigious methodological journals in statistics, for couple of years. Take a look at the content of the book and you will see that it covers almost everything that a researcher requires to know about statistics. I would summarize this book as what is useful in statistics to learn, with focus on methodology. This book is difficult to use as a textbook, and much more suitable as a reference. I can even call it the bible of statistics. The fundamental ideas in statistics are clearly explained, and argued with excellent motivating examples. Exercises are even better and completes the lecture, but they are not trivial at all. Some of the exercises are simpler versions of a research problem, being borrowed from a research article. You need some knowledge of mathematics in the level of calculus-based probability and statistics to start reading this text. After finishing the book you will have a general picture of what is statistics.

Superb - deeply interesting, beautifully written & fun to read. Some of this is reminding me what I learned at Cambridge 30+ years ago, much of it is new, all is brilliantly communicated. There are delightful touches, such as mini-biographical sketches in the margin of the mathematicians named in the text. The aphorism: "Nowadays it is very easy to make graphs. Unfortunately it is even easier to make bad graphs" should be dunned into everyone. Another gem: (p62) "I am uncertain about what will happen when I next roll a die, about the exact amount of money at present in my bank account, about the weather tomorrow, and about what will happen when I die. Does uncertainty mean the same thing in all these contexts? For which is variation due to repeated sampling meaningful, do you think?" And who could resist the shipwrecked statistician (p88) who, "marooned on a desert island...decides to realise his lifelong ambition of memorizing values of the normal integral; he hopes to make himself more attractive to the statisticienne of his dreams. His statistical tables have been ruined by salt water, but washed up on the beach he finds a solar-powered calculator..."?

This book is a real piece of art. The content, which is almost infinite, is always very simply presented and relevant. The theory is concise though it is efficient and straightforward. However one mustn't expect the book to be pedagogical. It is filled with numerous (and diverse! From the Darwin's experimentations data to the FTSE) examples which are continued and enriched throughout the text but they usually lack details and fail to emphasise the main definitions and topics. Some of the most important subjects of statistics are presented in half a page and many definitions which would have required a whole chapter hardly appear in italic here. The experienced student might be tempted to smile at this: "Given the presentation, are we really supposed to realise how important this is ?". This is the main reason why I wouldn't recommend this book to students looking for an introduction to statistics. However, the text is very agreeable to a self confident student who would like to have an overview of most the statistical models we use today. Plus, the diversity of examples makes it a good source of inspiration for the future statisticians looking a range for a topic to tackle. Even though, it sometimes seems hard to follow all the material, reading (and working on the exercices of) this book shall help the experienced reader getting a better understanding of statistics as a whole and discovering its depth, which is quite unique. Thank you, mister Davison ! P-S : sadly even if you pay or the hardback edition, you might get one of these "print on demand" books with very very poor quality compared to the original print.

E' un libro molto vasto che tocca una gran quantità di modelli statistici. Evidentemente non può approfondirli tutti con grande dettaglio ma è un'utilissima guida di riferimento. Il costo, poi, così contenuto, lo rende molto interessante per la propria biblioteca personale.