What do you think?
Rate this book
Why: A Guide to Finding and Using Causes
Can drinking coffee help people live longer? What makes a stock’s price go up? Why did you get the flu? Causal questions like these arise on a regular basis, but most people likely have not thought deeply about how to answer them. This book helps you think about causality in a structured What is a cause, what are causes good for, and what is compelling evidence of causality? Author Samantha Kleinberg shows you how to develop a set of tools for thinking more critically about causes. You’ll learn how to question claims, identify causes, make decisions based on causal information, and verify causes through further tests. Whether it’s figuring out what data you need, or understanding that the way you collect and prepare data affects the conclusions you can draw from it, Why will help you sharpen your causal inference skills.
284 pages, Paperback
Published January 5, 2016
28 people are currently reading
263 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 14 of 14 reviews
February 11, 2019
I rented it from the library, and it expired before I read the whole thing. I suppose the fact that I didn't even finish it is not in the books favor. It was ok.
March 27, 2016
Intuitively, we think we understand what constitutes a cause of a particular outcome and how we would go about finding causes to explain an observed event. But when you think you have a defined a general principle about causality, you immediately are confronted with exceptions. Early on, Kleinberg proposes a good, even if imperfect, definition of a cause: "something that makes an effect more likely, without which an effect would or could not occur, or that is capable of producing an effect under the right circumstances." Kleinberg then explains how psychology, time, correlation, observation, experimentation, all impact the determination of what causes a particular outcome. It's a worthwhile, nontechnical introduction to a complex topic.
March 28, 2020
I had read some of the reviews on this book here, and I must admit that I think they *caused* not an insignificant amount of scepticism as I read this book. And to be honest, mid-way to almost fully through the book, I thought it was a 3-star at best.
However, after reading the last chapter I realised that it might just be that I was treating this book as something it was not: a manual on how to find causes. Because despite what the sub-title of the book says, it's not really "a guide to finding and using causes".
To me, it's more useful to say that book is a guide to *understanding* causality, current methods to determine it, and limitations of these methods.
I found the last chapter sufficient cause to give this book four stars, because it was the chapter that finally opened my eyes to what this book was really about. It highlighted succinctly the difficulties of determining causality, the need for trade-offs, the need for multiple methods, and how sometimes despite our best intentions and willingness to pour resources into determining causality, we will fall short.
As somebody who works closely with data and determining causality, I cannot help but find relief in knowing that somebody far more experienced and an expert in causality might face the same difficulties I do; to know that when somebody in my organisation tells me to find the "impact" of something, or the "cause" of something, and the best I can do is provide probabilities causes or that it is nigh impossible, that it's not (necessarily) because I'm incompetent or lazy.
Worth a read if you're not simply looking for practical methods to determine causality. And like I mentioned, highly recommend this book simply for the last chapter alone.
However, after reading the last chapter I realised that it might just be that I was treating this book as something it was not: a manual on how to find causes. Because despite what the sub-title of the book says, it's not really "a guide to finding and using causes".
To me, it's more useful to say that book is a guide to *understanding* causality, current methods to determine it, and limitations of these methods.
I found the last chapter sufficient cause to give this book four stars, because it was the chapter that finally opened my eyes to what this book was really about. It highlighted succinctly the difficulties of determining causality, the need for trade-offs, the need for multiple methods, and how sometimes despite our best intentions and willingness to pour resources into determining causality, we will fall short.
As somebody who works closely with data and determining causality, I cannot help but find relief in knowing that somebody far more experienced and an expert in causality might face the same difficulties I do; to know that when somebody in my organisation tells me to find the "impact" of something, or the "cause" of something, and the best I can do is provide probabilities causes or that it is nigh impossible, that it's not (necessarily) because I'm incompetent or lazy.
Worth a read if you're not simply looking for practical methods to determine causality. And like I mentioned, highly recommend this book simply for the last chapter alone.
April 5, 2018
It's technical, but it's not a textbook. Provides a very good rundown of causal inference from both a philosophy and a stats perspective. Accessible to lay readers, but has lots of pointers to more detailed information for implementing the approaches it explains.
September 9, 2018
Unclear. The book is like a thesis. It fails to generate the reader's interest in the matter as the author is too fixed on the technicality of statistics rather than the cases happening/happened around the world.
July 2, 2025
Книга больше теоретическая с множеством теорий, методов и их хорошим разжевыванием. Слишком много деталей и экспериментов, но объяснения на аналогиях хорошие, понятные. Самая полезная часть вначале книги, а дальше идет излишнее детализирование, что напрочь отбивает желание читать.
August 25, 2018
Нельзя на все смотрят с одной стороны. Надо выбирать разные подходы. Всё имеет взаимосвязь, даже отдаленные изменения могут привести к глобальным следствиям и не всегда все это можно учесть
February 6, 2020
Correlation for prediction, causality for intervention.
November 28, 2021
Wonderful! Especially for lawyers!
September 15, 2022
This book is getting another site of understanding why is something happening.
This book make me to restructure my eye vision at another angle.
This book make me to restructure my eye vision at another angle.
August 23, 2016
The phrase "correlation is not causation" is one of the most famous phrases in the world of statistics, and science: anyone who did a bit amount of reading will complete the sentence before you can finish saying the second word.
But if correlation is not causation, then what on earth is causation? Or, is it even possible to have causation without correlation? The questions sound dead simple, yet they have such depth and importance that researchers are still trying to come up with explanations and techniques to clarify them. From trying to find whether a medication is really effective, to illuminating mysterious criminal cases, from trying to make predictions of future events based on past time series to making policy decisions for societies in order to get the intended effect, "finding and using causes" is an essential aspect, full of intellectual pitfalls.
One has to tread carefully in this area, and this book can be considered a good introduction to the topic. It is intended for people who are not experts on causality; avoiding rigorous technical explanations from the relevant fields, preferring to motivate the intuition of the reader by giving many real-world examples, some of them pretty surprising. It is unfortunate that it sometimes becomes repetitive and wordy, a good editor would probably be able to cut down 20% of the material without loss of the critical information and messages it conveys.
After reading the book, if you developed a more critical view towards popular science reporting such as "X therapy is found to cause Y", "politicians want to enact X so that society will be better off", etc., then I think the book will have achieved its goal. However, to go beyond that, you'll have to follow the contents of the extensive notes and references provided by the author, or read one of her more technical book "Causality, Probability, and Time".
But if correlation is not causation, then what on earth is causation? Or, is it even possible to have causation without correlation? The questions sound dead simple, yet they have such depth and importance that researchers are still trying to come up with explanations and techniques to clarify them. From trying to find whether a medication is really effective, to illuminating mysterious criminal cases, from trying to make predictions of future events based on past time series to making policy decisions for societies in order to get the intended effect, "finding and using causes" is an essential aspect, full of intellectual pitfalls.
One has to tread carefully in this area, and this book can be considered a good introduction to the topic. It is intended for people who are not experts on causality; avoiding rigorous technical explanations from the relevant fields, preferring to motivate the intuition of the reader by giving many real-world examples, some of them pretty surprising. It is unfortunate that it sometimes becomes repetitive and wordy, a good editor would probably be able to cut down 20% of the material without loss of the critical information and messages it conveys.
After reading the book, if you developed a more critical view towards popular science reporting such as "X therapy is found to cause Y", "politicians want to enact X so that society will be better off", etc., then I think the book will have achieved its goal. However, to go beyond that, you'll have to follow the contents of the extensive notes and references provided by the author, or read one of her more technical book "Causality, Probability, and Time".
May 4, 2016
En hel del matnyttigt, men den är grymt invecklad - som att läsa programmeringskod.
February 7, 2016
Useful
Useful book to understand correlation is not causation. The reference stories are too lengthy at best but the relevance of the academic components are good
Useful book to understand correlation is not causation. The reference stories are too lengthy at best but the relevance of the academic components are good
April 3, 2016
Really great book on finding causes. Completely changed my orientation of causes and effects.
Displaying 1 - 14 of 14 reviews