PENGUIN The Art of Statistics: Learning from Data

Essenzialmente non è un libro di statistica, ma un libro sulla statistica. Aiuta a comprendere concetti fondamentali troppo spesso trascurati nella usuale didattica “meccanicistica”. È vero che si “impara facendo”, ma se non sai quello che fai… Indispensabile affiancamento ai corsi tradizionali di statistica.

Intéressant mais pas génial. J’aime toujours apprendre. Mais ici je ne suis pas certain que le livre est vraiment utile. Sincèrement.

I very much enjoyed reading this book- cover to cover- the 380 pages of content- and then the additional glossary which in itself is a great resource. This book is not the "nuts and bolts" of how you do the standard statistical tests. This book assumes you have looked at other resources for the formulas. What this book does better than ANYTHING I have ever come across (and I have come across a lot of books on statistics and analytics over the years) is to get you to understand-really understand-- the key concepts, principles, assumptions and even mindsets that underlie the use of the basic-- and foundational-- statistical concepts. Prof Spiegelhalter has a remarkable ability to write and tell stories in clear and simple ways that get you to understand the essence of the principles. He walks you through many real and relevant examples so you understand the essentials of the underlying math models and mental models, and in parallel, that you also understand common mistakes and misconceptions. Some of you might say-- why do I need to visit-- or more likely- revisit these basic-- and ubiquitously applied- principles and methods? After all, isn't this becoming increasingly automated? Can't I just "trust" the outputs of my AI-enabled analytic systems? Exactly because the use of statistics has become so widespread, so deeply embedded in so many analytics and prediction systems-- it could never be more important than now to understand why people (and the models and systems they design) SO OFTEN misunderstand and misapply the basics. And these basics are foundational. They underlie so many aspects of modern analytics--- and even modern AI-based systems ultimately end up using these basics. All I can say is--- even though I have a MSc in Stats from Carnegie Mellon (1981), and have been involved with the use of data, statistics, analytics, and AI applications for decades-- I found Prof Spiegelhalter's way of walking a reader through the essentials of statistical thinking to be a joy to read. It has tremendously helped me to clarify my own understanding and mental models per statistical basics. This book truly works for people at ANY level. If you are just beginning with statistics- and you know the formulas but you still don't have the intuitions and insights- please read this book. If you work with stats on a regular basis--- EITHER because you are part of a team that uses statistical methods in your work products--- OR--- you regularly REVIEW work products that provide the results of analytic models-- You will find this book invaluable. And if you have to get other people to understand how to apply and how to NOT misapply basic statistical principles- this book is a hugely useful resource. Thank you Prof Spiegelhalter for producing this work. It is so easy to read. So easy to understand. And at the same time, so insightful, and so thoughtful-- in so many practical ways.

This book is incredibly engaging—I find it hard to put down and always want to keep it close. Through a series of diverse stories, the author skillfully demonstrates how statistics and machine learning are applied in everyday life. This storytelling approach makes the material both accessible and enjoyable, even for readers without a technical background. The inclusion of a glossary and references at the end of the book further supports non-experts, helping them follow along and deepen their understanding of key concepts.

My daughter, who is a mechanical engineer, loves this book and found it well written and organized