What do you think?
Rate this book
The Data Detective: Ten Easy Rules to Make Sense of Statistics
Today we think statistics are the enemy, numbers used to mislead and confuse us. That’s a mistake, Tim Harford says in The Data Detective. We shouldn’t be suspicious of statistics—we need to understand what they mean and how they can improve our lives: they are, at heart, human behavior seen through the prism of numbers and are often “the only way of grasping much of what is going on around us.” If we can toss aside our fears and learn to approach them clearly—understanding how our own preconceptions lead us astray—statistics can point to ways we can live better and work smarter.
As “perhaps the best popular economics writer in the world” (New Statesman), Tim Harford is an expert at taking complicated ideas and untangling them for millions of readers. In The Data Detective, he uses new research in science and psychology to set out ten strategies for using statistics to erase our biases and replace them with new ideas that use virtues like patience, curiosity, and good sense to better understand ourselves and the world. As a result, The Data Detective is a big-idea book about statistics and human behavior that is fresh, unexpected, and insightful.
As “perhaps the best popular economics writer in the world” (New Statesman), Tim Harford is an expert at taking complicated ideas and untangling them for millions of readers. In The Data Detective, he uses new research in science and psychology to set out ten strategies for using statistics to erase our biases and replace them with new ideas that use virtues like patience, curiosity, and good sense to better understand ourselves and the world. As a result, The Data Detective is a big-idea book about statistics and human behavior that is fresh, unexpected, and insightful.
334 pages, Kindle Edition
First published January 1, 2020
About the author
Tim Harford
41 books1,765 followersTim Harford is a member of the Financial Times editorial board. His column, “The Undercover Economist”, which reveals the economic ideas behind everyday experiences, is published in the Financial Times and syndicated around the world. He is also the only economist in the world to run a problem page, “Dear Economist”, in which FT readers’ personal problems are answered tongue-in-cheek with the latest economic theory.
--from the author's website
--from the author's website
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 30 of 726 reviews
Read
August 1, 2022Very readable overview of how to look at statistics (not in depth analysis, more your attitude towards what you read and questions you should ask yourself). Thoughtful, useful, and with a bunch of good stories.
Here noting my profound irritation that Florence Nightingale is remembered as a loving nurturing Lady with the Lamp bringing a kind word to wounded soldiers' bedsides. The woman was a *statistician*, damn it. She literally ran the numbers, put the data in an effective format, and did a marketing campaign to force a government policy change, and people have rebranded her as holding patients' hands. Bah.
Good read, with some excellent jokes.
Here noting my profound irritation that Florence Nightingale is remembered as a loving nurturing Lady with the Lamp bringing a kind word to wounded soldiers' bedsides. The woman was a *statistician*, damn it. She literally ran the numbers, put the data in an effective format, and did a marketing campaign to force a government policy change, and people have rebranded her as holding patients' hands. Bah.
Good read, with some excellent jokes.
October 6, 2020
This book could be called: "How Not To Be Fooled"
We all know the kind of person who says, "Oh, you can prove anything with statistics," and they suppose that cynicism makes them seem smart. In fact it's as dumb as saying, "It's possible to lie, so you should never believe anything." What Tim Harford does here is show you how not to be lied to. He provides a simple checklist to let you see what is really going on, so you can see past the tricks that politicians and pundits might use to bamboozle you.
Within a very short time you see how easy it is to think for yourself. You'll take joy in discovering the truth that is waiting there in the data. You'll recognize the obfuscations so that, presented with a set of facts, you can see for yourself what's really going on.
Guided by the example of statistical heroes like Florence Nightingale and Hans Rosling, you'll discover the power of being able to use data to understand the underlying patterns of the world. Sounds like what the Ancient One promised Dr Strange? Being a data detective is just as wonderful, but the "magic" at work is the power of reason and clear-sightedness that is within the grasp of us all. And the best bit? This book makes it fun!
We all know the kind of person who says, "Oh, you can prove anything with statistics," and they suppose that cynicism makes them seem smart. In fact it's as dumb as saying, "It's possible to lie, so you should never believe anything." What Tim Harford does here is show you how not to be lied to. He provides a simple checklist to let you see what is really going on, so you can see past the tricks that politicians and pundits might use to bamboozle you.
Within a very short time you see how easy it is to think for yourself. You'll take joy in discovering the truth that is waiting there in the data. You'll recognize the obfuscations so that, presented with a set of facts, you can see for yourself what's really going on.
Guided by the example of statistical heroes like Florence Nightingale and Hans Rosling, you'll discover the power of being able to use data to understand the underlying patterns of the world. Sounds like what the Ancient One promised Dr Strange? Being a data detective is just as wonderful, but the "magic" at work is the power of reason and clear-sightedness that is within the grasp of us all. And the best bit? This book makes it fun!
November 30, 2020
How to Make the World Add Up: Ten Rules for Thinking Differently About Numbers (2020) by Tim Harford is a book about how to interpret the myriad of statistics that now abound in the news. Harford has written for the Financial Times and hosts the excellent BBC Podcast on statistics in current Affairs, “More or Less”. So if you’re a ‘loyal listener’ you have a fair idea of what’s coming.
The rules are:
First, we should learn to stop and notice our emotional reaction to a claim, rather than accepting or rejecting it because of how it makes us feel.
Second, we should look for ways to combine the ‘bird’s eye’ statistical perspective with the ‘worm’s eye’ view from personal experience.
Third, we should look at the labels on the data we’re being given, and ask if we understand what’s really being described.
Fourth, we should look for comparisons and context, putting any claim into perspective.
Fifth, we should look behind the statistics at where they came from – and what other data might have vanished into obscurity.
Sixth, we should ask who is missing from the data we’re being shown, and whether our conclusions might differ if they were included.
Seventh, we should ask tough questions about algorithms and the big datasets that drive them, recognising that without intelligent openness they cannot be trusted.
Eighth, we should pay more attention to the bedrock of official statistics – and the sometimes heroic statisticians who protect it.
Ninth, we should look under the surface of any beautiful graph or chart.
And tenth, we should keep an open mind, asking how we might be mistaken, and whether the facts have changed.
These rules are all well though through and each chapter has a narrative and a good number of interesting tales. Harford writes really well and knows his subject intimately which tends to result in a good book.
Nicely the book also returns to ‘How to Lie with Statistics’ and Harford describes that statistics when used at their best are the absolute opposite, they enable us to better understand the world if we careful and curious with their use.
How to Make the World Add Up is another excellent book from an accomplished author. It’s definitely worth a read for anyone interested in how to read and comprehend statistics.
The rules are:
First, we should learn to stop and notice our emotional reaction to a claim, rather than accepting or rejecting it because of how it makes us feel.
Second, we should look for ways to combine the ‘bird’s eye’ statistical perspective with the ‘worm’s eye’ view from personal experience.
Third, we should look at the labels on the data we’re being given, and ask if we understand what’s really being described.
Fourth, we should look for comparisons and context, putting any claim into perspective.
Fifth, we should look behind the statistics at where they came from – and what other data might have vanished into obscurity.
Sixth, we should ask who is missing from the data we’re being shown, and whether our conclusions might differ if they were included.
Seventh, we should ask tough questions about algorithms and the big datasets that drive them, recognising that without intelligent openness they cannot be trusted.
Eighth, we should pay more attention to the bedrock of official statistics – and the sometimes heroic statisticians who protect it.
Ninth, we should look under the surface of any beautiful graph or chart.
And tenth, we should keep an open mind, asking how we might be mistaken, and whether the facts have changed.
These rules are all well though through and each chapter has a narrative and a good number of interesting tales. Harford writes really well and knows his subject intimately which tends to result in a good book.
Nicely the book also returns to ‘How to Lie with Statistics’ and Harford describes that statistics when used at their best are the absolute opposite, they enable us to better understand the world if we careful and curious with their use.
How to Make the World Add Up is another excellent book from an accomplished author. It’s definitely worth a read for anyone interested in how to read and comprehend statistics.
September 5, 2022
I like Tim Harford's podcasts and previous writing, but this fell flat for me. There's nothing wrong overall with the underlying content about statistics, and that's something to cheer for. But science nerds will likely know that stuff already. The plus-value I was hoping for was something on how to get factual information through to people who seem resistant to data. And here, I'm afraid to say Harford's conclusion is ultimately disappointing.
Harford gives examples of big life-saving behavior changes based on data, like the Doll and Hill epidemiology on lung cancer leading to tobacco control, and Florence Nightingale's graphs leading to sanitary reforms. But those changes did not happen because the average citizen or even the average doctor ever read the original research. So there's a disconnect between those examples and the general thrust of this book of trying to get everyone to scrutinize data.
...So maybe that's the point the book is making unwittingly. If you're not willing to dig deep and know what you are talking about, then please just stop talking.
Harford gives examples of big life-saving behavior changes based on data, like the Doll and Hill epidemiology on lung cancer leading to tobacco control, and Florence Nightingale's graphs leading to sanitary reforms. But those changes did not happen because the average citizen or even the average doctor ever read the original research. So there's a disconnect between those examples and the general thrust of this book of trying to get everyone to scrutinize data.
...So maybe that's the point the book is making unwittingly. If you're not willing to dig deep and know what you are talking about, then please just stop talking.
December 21, 2022
Briefly - Fascinating, thought provoking and made me more curious.
In full
There is a good opening here on statistics generally which I found helpful. While this books was originally published a little while ago it does appear to have been updated since the previous edition with quite a bit about Covid and statistics. Over the course of the book Tim Harford offers 10 "rules" for thinking about the numbers we see presented to us. Each rule has a chapter and the topic of the rule is discussed with pertinent illustrations using real life statistics.
The range of examples in this book is quite remarkable really. While thinking about this review I reflected on the fact that, while reading this, I learnt about both algorithms and the work of Florence Nightingale. I doubt that can be said about many books! Each chapter offers statistical examples and then considers how an ordinary person could make more sense of the information. Equally the author suggests ways we might test the veracity of the information we are given. Attention is paid to simple but potentially misleading words such as what is actually meant by "everyone" or "all - very interesting.
Tim Harford suggests we consider our own experiences to some degree at leat. An example of this would be the Transport for London statistics on average occupancy of tube trains and buses. The author feels from his own experience that he has never travelled on an "averagely" occupied service! This allows him to look at the issues with averages as well as how the data may have been collected.
Returning to algorithms I confess I was not aware that there had ever been a "Google Flu predictor" that appeared to be very accurate at predicting outbreaks of flu. After a while it ceased to be accurate. In a nutshell - no one actually knew why it worked! - beware algorithms that people don't actually understand is the message.
For me the whole book was extremely readable. Some chapters appealed more to me than others but that is inevitable. One that has stuck with me is the chapter than has information on the work of Florence Nightingale. I confess quite a bit of it came as a surprise to me. The chapter is focussed on infographics or how statistics can be presented. As someone who has from time to time presented graphs to illustrate information quite a bit of this hit home to me. The use and misuse of infographics is fascinating.
I've been a "fan" of Tim Harford for some years now due to his programmes on Radio 4. I always found them interesting and accessible - this book is very similar. Assuming you have any interest in the subject it makes for a very easy read. However it is also thought provoking and interesting. Above all Tim Harford urges the reader to "be curious" - I would suggest that anyone who is will thoroughly enjoy this book.
Note - I received an advance digital copy of this book from the publisher in exchange for a fair review
In full
There is a good opening here on statistics generally which I found helpful. While this books was originally published a little while ago it does appear to have been updated since the previous edition with quite a bit about Covid and statistics. Over the course of the book Tim Harford offers 10 "rules" for thinking about the numbers we see presented to us. Each rule has a chapter and the topic of the rule is discussed with pertinent illustrations using real life statistics.
The range of examples in this book is quite remarkable really. While thinking about this review I reflected on the fact that, while reading this, I learnt about both algorithms and the work of Florence Nightingale. I doubt that can be said about many books! Each chapter offers statistical examples and then considers how an ordinary person could make more sense of the information. Equally the author suggests ways we might test the veracity of the information we are given. Attention is paid to simple but potentially misleading words such as what is actually meant by "everyone" or "all - very interesting.
Tim Harford suggests we consider our own experiences to some degree at leat. An example of this would be the Transport for London statistics on average occupancy of tube trains and buses. The author feels from his own experience that he has never travelled on an "averagely" occupied service! This allows him to look at the issues with averages as well as how the data may have been collected.
Returning to algorithms I confess I was not aware that there had ever been a "Google Flu predictor" that appeared to be very accurate at predicting outbreaks of flu. After a while it ceased to be accurate. In a nutshell - no one actually knew why it worked! - beware algorithms that people don't actually understand is the message.
For me the whole book was extremely readable. Some chapters appealed more to me than others but that is inevitable. One that has stuck with me is the chapter than has information on the work of Florence Nightingale. I confess quite a bit of it came as a surprise to me. The chapter is focussed on infographics or how statistics can be presented. As someone who has from time to time presented graphs to illustrate information quite a bit of this hit home to me. The use and misuse of infographics is fascinating.
I've been a "fan" of Tim Harford for some years now due to his programmes on Radio 4. I always found them interesting and accessible - this book is very similar. Assuming you have any interest in the subject it makes for a very easy read. However it is also thought provoking and interesting. Above all Tim Harford urges the reader to "be curious" - I would suggest that anyone who is will thoroughly enjoy this book.
Note - I received an advance digital copy of this book from the publisher in exchange for a fair review
May 26, 2022
This book was written by an economist who is also a columnist & broadcaster, not to mention an honorary fellow of The Royal Statistical Society. No wonder this book, basically about statistics, is well written, empathetic and full of interesting examples.
It consists of 10 rules telling how to think critically when you see statistics (figures, studies, charts). The good news none was technical (you know, p-value or mean average vs median), everything was beyond technical. The rules were about being careful about our emotion, being aware of the backstory, being aware of who was included and who was not in the data, and common advice that Adam Grant & Daniel Kahneman would give you: keep an open mind.
If you like Hans Rosling's 'Factfulness', you will like this book. However, I must say that this will also bring you trust issues; for scientific studies and even for statistics published by national authority 😂 I was a great believer of Iyengar & Lepper jam study about how choices can be motivating. Now, I don't know whom to really believe anymore 😂😂😂 Also it revealed how a country could actually manipulate the statistics of inflation and lie to the rest of the world.
Well, if you're interested in critical thinking and statistics, I heavily recommend this one.
It consists of 10 rules telling how to think critically when you see statistics (figures, studies, charts). The good news none was technical (you know, p-value or mean average vs median), everything was beyond technical. The rules were about being careful about our emotion, being aware of the backstory, being aware of who was included and who was not in the data, and common advice that Adam Grant & Daniel Kahneman would give you: keep an open mind.
If you like Hans Rosling's 'Factfulness', you will like this book. However, I must say that this will also bring you trust issues; for scientific studies and even for statistics published by national authority 😂 I was a great believer of Iyengar & Lepper jam study about how choices can be motivating. Now, I don't know whom to really believe anymore 😂😂😂 Also it revealed how a country could actually manipulate the statistics of inflation and lie to the rest of the world.
Well, if you're interested in critical thinking and statistics, I heavily recommend this one.
March 18, 2021
الإحصاء نافذة على على الواقع. صورة رقمية تدّعي الدقة والمصداقية. ولأنّ الصورة الإحصائية معادلاتٌ رياضية ونواتج رقمية، فإنّها تبدو مهيبة! يبتعد الناس عن مساءلتها، استسلامًا لها، وهيبةً لمعادلاتها، فلا يطيلون التحديق فيها.
لا يهدف "تيم هارفورد" إلى إقناعنا بقدرة الإحصاء على خداعنا، ولا يطلب منا ترك الإحصاءات والزهد بها. لكنّه يقدم لنا كتابًا يعيننا على النظر بعمق إلى كلّ رقم أو رسمٍ إحصائي يُقدّم إلينا، أو يتم الترويج له والبناء عليه.
مرةً يشير الكاتب إلى الخلل في المتلقي، ومرةً يشير إلى الخلل في المُخرج الإحصائي، ومرةً إلى إشكاليات النموذج التقني نفسه.
في عشرة فصول؛ يُقدّم "هارفورد" عشر نصائح لقرّائه. مستعينًا بالأمثلة القصصية المطولة، وبالإشارة إلى دراسات واقتباسات فكرية من هنا وهناك. ويختم كتابه بوصية جامعة نقاوم فيها بريق الرقم أو الرسم الإحصائي، أنْ كونوا فضوليين.
كتابٌ جيّد ونصائح مهمة في زمنٍ رقمي بامتياز، وأمثلة كاشفة. وربما يجد فيه القارئ أمثلة يحياها في سياق جائحة كورونا، التي لم يفت الكاتب الاستشهاد ببعض أحداثها.
ملاحظتي السلبية عن الكتاب تتمحور حول هذا الأسلوب الذي تصطبغ به كتابات "تيم هارفورد". أقصد الأسلوب المغرق في القصصية، إسهابًا واستطرادًا، يقابله ميلي الشخصي للتجريد والاختصار.
لا يهدف "تيم هارفورد" إلى إقناعنا بقدرة الإحصاء على خداعنا، ولا يطلب منا ترك الإحصاءات والزهد بها. لكنّه يقدم لنا كتابًا يعيننا على النظر بعمق إلى كلّ رقم أو رسمٍ إحصائي يُقدّم إلينا، أو يتم الترويج له والبناء عليه.
مرةً يشير الكاتب إلى الخلل في المتلقي، ومرةً يشير إلى الخلل في المُخرج الإحصائي، ومرةً إلى إشكاليات النموذج التقني نفسه.
في عشرة فصول؛ يُقدّم "هارفورد" عشر نصائح لقرّائه. مستعينًا بالأمثلة القصصية المطولة، وبالإشارة إلى دراسات واقتباسات فكرية من هنا وهناك. ويختم كتابه بوصية جامعة نقاوم فيها بريق الرقم أو الرسم الإحصائي، أنْ كونوا فضوليين.
كتابٌ جيّد ونصائح مهمة في زمنٍ رقمي بامتياز، وأمثلة كاشفة. وربما يجد فيه القارئ أمثلة يحياها في سياق جائحة كورونا، التي لم يفت الكاتب الاستشهاد ببعض أحداثها.
ملاحظتي السلبية عن الكتاب تتمحور حول هذا الأسلوب الذي تصطبغ به كتابات "تيم هارفورد". أقصد الأسلوب المغرق في القصصية، إسهابًا واستطرادًا، يقابله ميلي الشخصي للتجريد والاختصار.
June 1, 2022
I love my science books and I read this one at a perfect time. I have not had much brain power left over to use for my pleasure reading so non-fiction has been a struggle. I had enough of a grasp of this topic that I could check out of the parts I mostly knew. Some great real world examples to emphasize his points, I think Hartford's biggest take home for me was to examine how a certain statistic printed in an articly or paper, affects me emotionally, before having it affect me intellectually.
I also appreciated his message that although statistics can and often will be used to bend truth or even lie outright, they and the data behind it should be appreciated for the insights they bring.
I also appreciated his message that although statistics can and often will be used to bend truth or even lie outright, they and the data behind it should be appreciated for the insights they bring.
February 4, 2021
How to see the Truth with Statistics – illustrated with good stories
Darrell Huff's classic book, How to Lie with Statistics, warns us not to get misled by statistics. In this book, Tim Harford tells us how to see the Truth with statistics. He gives us ten rules of thumb for thinking about reported numbers. But the real reason to read the book is for all of the stories Harford tells to illustrate each rule. It's like the greatest hits from the BBC radio show he has hosted for years.
Here is a simple summary of the ten rules:
1. When considering new information, pay attention to how it makes you feel. Your emotions can influence you to dismiss accurate statistics that you do not like and to embrace false statistics that you do like
2. Sometimes your personal experience (a worm's eye view) conflicts with a bird's-eye view statistic. For example, the subway may be only half full on average during the day but packed every time you ride it (during rush hour). Both perspectives help you understand the truth.
3. Make sure you understand what is being counted. When counting beans, the definition of a bean matters.
4. Look for information that can put a statistic into context, like the trend, the scale, or how it compares to other situations.
5. Try to learn where the statistics came from (the backstory) – and what other data might have vanished into obscurity.
6. Ask who is missing from the data, and would our conclusions differ if they were included.
7. Ask tough questions about algorithms and the big datasets that drive them, recognizing that without intelligent openness they cannot be trusted.
8. Pay more attention to the bedrock of official statistics – and the sometimes heroic statisticians who protect it.
9. Look under the surface of any beautiful graph or chart. Don't let the beauty mislead you.
10. Keep an open mind, asking how we might be mistaken and whether the facts have changed.
Those ten tips sound boring, but Harford's stories are not. Most of them show the power of useful statistics. As Harford says: "Good statistics are not smoke and mirrors; in fact, they help us see more clearly. Good statistics are like a telescope for an astronomer, a microscope for a bacteriologist, or an X-ray for a radiologist."
Darrell Huff's classic book, How to Lie with Statistics, warns us not to get misled by statistics. In this book, Tim Harford tells us how to see the Truth with statistics. He gives us ten rules of thumb for thinking about reported numbers. But the real reason to read the book is for all of the stories Harford tells to illustrate each rule. It's like the greatest hits from the BBC radio show he has hosted for years.
Here is a simple summary of the ten rules:
1. When considering new information, pay attention to how it makes you feel. Your emotions can influence you to dismiss accurate statistics that you do not like and to embrace false statistics that you do like
2. Sometimes your personal experience (a worm's eye view) conflicts with a bird's-eye view statistic. For example, the subway may be only half full on average during the day but packed every time you ride it (during rush hour). Both perspectives help you understand the truth.
3. Make sure you understand what is being counted. When counting beans, the definition of a bean matters.
4. Look for information that can put a statistic into context, like the trend, the scale, or how it compares to other situations.
5. Try to learn where the statistics came from (the backstory) – and what other data might have vanished into obscurity.
6. Ask who is missing from the data, and would our conclusions differ if they were included.
7. Ask tough questions about algorithms and the big datasets that drive them, recognizing that without intelligent openness they cannot be trusted.
8. Pay more attention to the bedrock of official statistics – and the sometimes heroic statisticians who protect it.
9. Look under the surface of any beautiful graph or chart. Don't let the beauty mislead you.
10. Keep an open mind, asking how we might be mistaken and whether the facts have changed.
Those ten tips sound boring, but Harford's stories are not. Most of them show the power of useful statistics. As Harford says: "Good statistics are not smoke and mirrors; in fact, they help us see more clearly. Good statistics are like a telescope for an astronomer, a microscope for a bacteriologist, or an X-ray for a radiologist."
October 9, 2020
This is a very timely book for our ‘post-truth’ world. The author acts like a kind of data hygienist, giving us advice on how to spot dodgy statistics, dodgy presentation of statistics and, equally importantly, our own prejudices. Wanting to believe a thing is true – or not true – is the biggest risk we face in interpreting stats well.
This is all well and good and worthwhile, but here’s what I really liked about this book: it’s passionate. A passionate statistics book may seem like an oxymoron, but here it is anyway. The chapter (Ch.8) on the near-heroic role played by statisticians at different times and in different places was particularly affecting. The author highlights the case of Andreas Georgiou, who left behind a dazzling career at the IMF to run Greece’s new statistical agency during the white heat of the Eurozone crisis. Georgiou started to bring rigour and credibility – and some very uncomfortable truths – into the inchoate world of Greek financial statistics. In return, the Greek state harassed him, repeatedly dragged him through the courts, and nearly imprisoned him. The blunt truth that this book points towards is that modern societies run on statistics: they run better on good statistics, and worse on bad ones.
As you might expect from a hyper-accomplished communicator like Harford, there is a gorgeous inter-leaving of fluent explanation with colourful anecdote throughout the book. The mad story in Chapter 1 of the Dutch master-forger of Dutch masters, who duped both the Nazis and the country’s foremost art critics, is worth the cover price alone. In the penultimate chapter on data visualisation, Florence Nightingale plays a star turn as she proves herself to be far, far more than a footnote in the development of statistical analysis. In the final chapter, the author focusses on John Maynard Keynes’ life and achievements, from which colourful anecdotes burst like a glitter bomb.
To repeat: this is an important book with a big heart. The title seems a bit off: it sounds like it belongs to the sub-genre of popular books on pure maths (which does exist). But that’s a tiny quibble. This is a book to read and relish.
This is all well and good and worthwhile, but here’s what I really liked about this book: it’s passionate. A passionate statistics book may seem like an oxymoron, but here it is anyway. The chapter (Ch.8) on the near-heroic role played by statisticians at different times and in different places was particularly affecting. The author highlights the case of Andreas Georgiou, who left behind a dazzling career at the IMF to run Greece’s new statistical agency during the white heat of the Eurozone crisis. Georgiou started to bring rigour and credibility – and some very uncomfortable truths – into the inchoate world of Greek financial statistics. In return, the Greek state harassed him, repeatedly dragged him through the courts, and nearly imprisoned him. The blunt truth that this book points towards is that modern societies run on statistics: they run better on good statistics, and worse on bad ones.
As you might expect from a hyper-accomplished communicator like Harford, there is a gorgeous inter-leaving of fluent explanation with colourful anecdote throughout the book. The mad story in Chapter 1 of the Dutch master-forger of Dutch masters, who duped both the Nazis and the country’s foremost art critics, is worth the cover price alone. In the penultimate chapter on data visualisation, Florence Nightingale plays a star turn as she proves herself to be far, far more than a footnote in the development of statistical analysis. In the final chapter, the author focusses on John Maynard Keynes’ life and achievements, from which colourful anecdotes burst like a glitter bomb.
To repeat: this is an important book with a big heart. The title seems a bit off: it sounds like it belongs to the sub-genre of popular books on pure maths (which does exist). But that’s a tiny quibble. This is a book to read and relish.
November 29, 2021
This book could have been easily made into a long essay. I expected the book to teach me statistical thinking or to better introduce statistics as a method, informally. It didn't. Instead, the author chose to include tens of examples of how people made mistakes, whether intentionally or unintentionally, while working with data. The ten rules are good reminders to not fall into mistakes, or to put it better: to avoid being fooled by other people's statistical arguments, but that's not what the title of the book suggests. It should have been titled: "How Not to Be Fooled by Data: Ten Key Questions".
June 18, 2021
I thought How To Make The World Add Up was excellent. I expected to read a chapter or two, take a break and come back to it as I often do with dense books about science or maths, but in fact I was hooked and read it with huge enjoyment from beginning to end.
Tim Harford’s message is that statistics are a vital tool in understanding the world, but that we need to be informed, thoughtful interpreters of what we are told. We must be aware of the way in which statistics and their presentation can be misleading, either deliberately or inadvertently, and also of our own prejudices and biases in how we receive and respond to what we hear and see. Just as one example, he points out that we often respond to a statistic which supports a belief with “Can I believe this?” but to one that apparently contradicts what we want to believe with “Do I have to believe this?” which leads to very different standards of rigour when we consider them.
It’s a very important and timely message. I love that Harford isn’t just trying to debunk bogus or misleading statistical claims (although he is very good at pointing out some tactics used by those wishing to distort or deceive), but emphasises the essential role good, solid statistical data and their analysis play in our lives. He gives us ten rules to apply when confronted with a statistic to try to decide on its veracity and usefulness. They are excellent, thoughtful rules which have deepened my understanding of the world, for which I am very grateful.
Tim Harford is an excellent communicator about statistics, as fellow Loyal Listeners to his Radio 4 programme, More Or Less, will know. I think he is even better in writing, partly because he has a chance to develop his engaging style a little more and partly because the humour is genuinely humorous, while it can feel a little laboured in the broadcasts. Whatever the reason, this is a pleasure to read; it is clear, thoughtful, witty, wise, balanced and very, very interesting. Very warmly recommended.
(My thanks to Bridge Street Press for an ARC via NetGalley.)
Tim Harford’s message is that statistics are a vital tool in understanding the world, but that we need to be informed, thoughtful interpreters of what we are told. We must be aware of the way in which statistics and their presentation can be misleading, either deliberately or inadvertently, and also of our own prejudices and biases in how we receive and respond to what we hear and see. Just as one example, he points out that we often respond to a statistic which supports a belief with “Can I believe this?” but to one that apparently contradicts what we want to believe with “Do I have to believe this?” which leads to very different standards of rigour when we consider them.
It’s a very important and timely message. I love that Harford isn’t just trying to debunk bogus or misleading statistical claims (although he is very good at pointing out some tactics used by those wishing to distort or deceive), but emphasises the essential role good, solid statistical data and their analysis play in our lives. He gives us ten rules to apply when confronted with a statistic to try to decide on its veracity and usefulness. They are excellent, thoughtful rules which have deepened my understanding of the world, for which I am very grateful.
Tim Harford is an excellent communicator about statistics, as fellow Loyal Listeners to his Radio 4 programme, More Or Less, will know. I think he is even better in writing, partly because he has a chance to develop his engaging style a little more and partly because the humour is genuinely humorous, while it can feel a little laboured in the broadcasts. Whatever the reason, this is a pleasure to read; it is clear, thoughtful, witty, wise, balanced and very, very interesting. Very warmly recommended.
(My thanks to Bridge Street Press for an ARC via NetGalley.)
September 3, 2022
Q:
Premature enumeration is an equal-opportunity blunder... (c)
Q:
... try to take both perspectives—the worm’s-eye view as well as the bird’s-eye view. They will usually show you something different, and they will sometimes pose a puzzle: How could both views be true? That should be the beginning of an investigation. (c)
Q:
It wasn’t a difference in reality, but a difference in how that reality was being recorded. (c)
Q:
Michael Blastland, co-creator of More or Less, imagines looking at two sheep in a field. How many sheep in the field? Two, of course. Except that one of the sheep isn’t a sheep, it’s a lamb. And the other sheep is heavily pregnant—in fact, she’s in labor, about to give birth at any moment. How many sheep again? One? Two? Two and a half? Counting to three just got difficult. (c)
Q:
It’s a frequent topic of conversation with my wife. The radio that sits on top of the refrigerator will carry some statistical claim into our home over breakfast—a political sound bite, or the dramatic conclusion of some research. For example, “A new study shows that children who play violent video games are more likely to be violent in reality.” Despite having known my limitations for twenty years, my wife can’t quite rid herself of the illusion that I have a huge spreadsheet in my head, full of every statistic in creation. So she will turn to me and ask, “Is that true?” Very occasionally I happen to have recently researched the issue and know the answer, but far more often I can only reply, “It all depends on what they mean . . .”
I’m not trying to model some radical philosophical skepticism—or annoy my wife. I’m just pointing out that I don’t fully understand what the claim means, so I am hardly in a position (yet) to know whether it might be true. For example, what is meant by a “violent video game”? Does Pac-Man count? Pac-Man commits heinous acts, notably swallowing sentient creatures alive. Or what about Space Invaders? There’s nothing to do in Space Invaders but shoot and avoid being shot. But perhaps that is not quite what the researchers meant. Until I know what they did mean, I don’t know much.
And how about “play”; what does that mean? Perhaps the researchers had children* fill in questionnaires to identify those who play violent games for many hours in a typical week. Or perhaps they recruited some experimental subjects to play a game for twenty minutes in a laboratory, then did some kind of test to see if they’d become more “violent in reality”—and how is that defined, anyway?
“Many studies won’t measure violence,” says Rebecca Goldin, a mathematician and director of the statistical literacy project STATS.6 “They’ll measure something else such as aggressive behavior.” And aggressive behavior itself is not easy to measure because it is not easy to define. One influential study of video games—I promise I’m not making this up—measured aggressive behavior by inviting people to add hot sauce to a drink that someone else would consume. This “hot sauce paradigm” was described as a “direct and unambiguous” assessment of aggression.7 I am not a social psychologist, so perhaps that’s reasonable. Perhaps. But clearly, like “baby” or “sheep” or “nurse,” apparently commonsense words such as “violent” and “play” can hide a lot of wiggle room. (c)
Q:
“Gun death” doesn’t sound like a complicated concept: a gun is a gun and dead is dead. Then again, nor does “sheep,” so we should pause to check our intuition. Even the year of death, 2017, isn’t as straightforward as you might think. For example, in the UK in 2016, the homicide rate rose sharply. This was because an official inquest finally ruled that ninety-six people who died in a crush at the Hillsborough soccer stadium in 1989 had been unlawfully killed. Initially seen as accidental, those deaths officially became homicides in 2016. This is an extreme example, but there are often delays between when somebody died and when the cause of death was officially registered. (c)
Q: Binge drinking seems very different from anorexia. ...
There is an enormous gulf between excessive exercise and killing yourself. (c)
Q:
Much of the data visualization that bombards us today is decoration at best, and distraction or even disinformation at worst. The decorative function is surprisingly common, perhaps because the data visualization teams of many media organizations are part of the art departments. They are led by people whose skills and experience are not in statistics but in illustration or graphic design. The emphasis is on the visualization, not on the data. It is, above all, a picture.
The most egregious examples of numbers as decoration are nothing more than the same old number in a large, striking font. (c)
Premature enumeration is an equal-opportunity blunder... (c)
Q:
... try to take both perspectives—the worm’s-eye view as well as the bird’s-eye view. They will usually show you something different, and they will sometimes pose a puzzle: How could both views be true? That should be the beginning of an investigation. (c)
Q:
It wasn’t a difference in reality, but a difference in how that reality was being recorded. (c)
Q:
Michael Blastland, co-creator of More or Less, imagines looking at two sheep in a field. How many sheep in the field? Two, of course. Except that one of the sheep isn’t a sheep, it’s a lamb. And the other sheep is heavily pregnant—in fact, she’s in labor, about to give birth at any moment. How many sheep again? One? Two? Two and a half? Counting to three just got difficult. (c)
Q:
It’s a frequent topic of conversation with my wife. The radio that sits on top of the refrigerator will carry some statistical claim into our home over breakfast—a political sound bite, or the dramatic conclusion of some research. For example, “A new study shows that children who play violent video games are more likely to be violent in reality.” Despite having known my limitations for twenty years, my wife can’t quite rid herself of the illusion that I have a huge spreadsheet in my head, full of every statistic in creation. So she will turn to me and ask, “Is that true?” Very occasionally I happen to have recently researched the issue and know the answer, but far more often I can only reply, “It all depends on what they mean . . .”
I’m not trying to model some radical philosophical skepticism—or annoy my wife. I’m just pointing out that I don’t fully understand what the claim means, so I am hardly in a position (yet) to know whether it might be true. For example, what is meant by a “violent video game”? Does Pac-Man count? Pac-Man commits heinous acts, notably swallowing sentient creatures alive. Or what about Space Invaders? There’s nothing to do in Space Invaders but shoot and avoid being shot. But perhaps that is not quite what the researchers meant. Until I know what they did mean, I don’t know much.
And how about “play”; what does that mean? Perhaps the researchers had children* fill in questionnaires to identify those who play violent games for many hours in a typical week. Or perhaps they recruited some experimental subjects to play a game for twenty minutes in a laboratory, then did some kind of test to see if they’d become more “violent in reality”—and how is that defined, anyway?
“Many studies won’t measure violence,” says Rebecca Goldin, a mathematician and director of the statistical literacy project STATS.6 “They’ll measure something else such as aggressive behavior.” And aggressive behavior itself is not easy to measure because it is not easy to define. One influential study of video games—I promise I’m not making this up—measured aggressive behavior by inviting people to add hot sauce to a drink that someone else would consume. This “hot sauce paradigm” was described as a “direct and unambiguous” assessment of aggression.7 I am not a social psychologist, so perhaps that’s reasonable. Perhaps. But clearly, like “baby” or “sheep” or “nurse,” apparently commonsense words such as “violent” and “play” can hide a lot of wiggle room. (c)
Q:
“Gun death” doesn’t sound like a complicated concept: a gun is a gun and dead is dead. Then again, nor does “sheep,” so we should pause to check our intuition. Even the year of death, 2017, isn’t as straightforward as you might think. For example, in the UK in 2016, the homicide rate rose sharply. This was because an official inquest finally ruled that ninety-six people who died in a crush at the Hillsborough soccer stadium in 1989 had been unlawfully killed. Initially seen as accidental, those deaths officially became homicides in 2016. This is an extreme example, but there are often delays between when somebody died and when the cause of death was officially registered. (c)
Q: Binge drinking seems very different from anorexia. ...
There is an enormous gulf between excessive exercise and killing yourself. (c)
Q:
Much of the data visualization that bombards us today is decoration at best, and distraction or even disinformation at worst. The decorative function is surprisingly common, perhaps because the data visualization teams of many media organizations are part of the art departments. They are led by people whose skills and experience are not in statistics but in illustration or graphic design. The emphasis is on the visualization, not on the data. It is, above all, a picture.
The most egregious examples of numbers as decoration are nothing more than the same old number in a large, striking font. (c)
February 20, 2021
Well, I suppose I should relate the "ten easy rules." 1. Learn to stop and notice one's emotional reaction to a claim, rather than accepting or rejecting it because of how it makes one feel. 2. Look for ways to combine the bird's-eye statistical perspective with the worm's-eye view from personal experience. 3. Look at the labels on the data one is given, asking if one understands what's really being described. 4. Look for comparison and context, putting any claim into perspective. 5. Look behind the statistics at their source--and at what other data might have vanished into obscurity. 6. Ask who is missing from the data being shown, and ask whether one's conclusions might differ if they were included. 7. Ask tough questions about algorithms and the big datasets that drive them, recognizing that without intelligent openness they cannot be trusted. 8. Pay more attention to the bedrock of official statistics. 9. Look under the surface of any eye-catching graph or chart. 10. Keep an open mind, asking how one might be mistaken, or whether the facts have changed.
Rather uncontroversial as far as they go. Harford's strength is in explanation and illustration. The book is basically a long way of teaching people to slow down and take statistics with a grain of salt. Inasmuch as that advice is heeded, the book should be a success.
Rather uncontroversial as far as they go. Harford's strength is in explanation and illustration. The book is basically a long way of teaching people to slow down and take statistics with a grain of salt. Inasmuch as that advice is heeded, the book should be a success.
August 22, 2021
This book didn't click with me and I'm not sure why.
Maybe it's because I've read a lot of this stuff before in different forms. Maybe a year and a half in a pandemic I'm tired of reading through virus statistics in the news and trying to make sense of it. Maybe I'm depressed about the state of the world!
...anyway, there were some interesting bits. Harford starts off by talking about the book "How to Lie with Statistics" by Darrell Huff, which was published in 1954 and became exceptionally famous for a book about statistics. (I had a copy growing up!) Reading it makes you more skeptical about numbers you see in the press.
But! Another thing that happened in 1954 was that one of the first studies that showed smoking cigarettes caused lung cancer came out. And we needed statistics to figure that out. And guess who testfied before Congress 11 years later on behalf of the cigarette companies, arguing that the studies were wrong and cigarettes were fine? The same Darrell Huff.
This is a powerful story! Harford's argument is that while we should be careful, we shouldn't be too cynical about statistics, because they can show us things that are almost impossible to see otherwise. The rest of the book is a list of rules to follow when you're looking at statistics to give yourself a better chance of understanding them:
1. Search your feelings - try to avoid emotional responses to data because it can cloud your judgment. (if you see a graph that you agree with, be extra careful before sharing it!)
2. Ponder your personal experience - if your personal experience contradicts some statistics you're seeing, maybe you should take a closer look at the statistics. Harford gives the example of measuring the average number of riders on a London Underground train throughout the day, which was pretty low. But his personal experience is that every time he takes a train to work it's extremely crowded. But these can both be true if most people are taking the train at the same time, and perhaps a better metric is the average number of other people on a train when a person gets on, or something like that. But conversely, lung cancer is rare enough that you can't trust your own intuition; you have to look at the statistics. I'm not real clear on how you're supposed to differentiate between these cases...
3. Avoid premature enumeration - be careful about what's actually being counted (for example, there were ~40,000 gun deaths in the US in 2017, but ~60% of these are suicides, which can definitely frame how you think about the issue differently)
4. Step back and enjoy the view - getting perspective is important, either by looking at different times, or asking "is that a big number?". Harford also points out that bad news tends to overpower good news in the media, and one reason for that is that bad news tends to happen suddenly, while good news happens slowly over time.
5. Get the backstory - it's common to see outlier successes without seeing all the common failures. (~40% of Kickstarter projects don't meet their goal, but if you hear about one in the media it's one that's been a huge success) Harford also talks about publication bias, and the fact that there's also a bias to publish novel and surprising discoveries but not replication attempts that might debunk previous "discoveries". One interesting point is that the "backfire effect" (where fact-checking can make people more likely to believe a false claim) is actually unusual and fact-checking does help in most cases. This chapter also talks about the Cochrane Collaboration, which is a group that looks at published medical papers and creates research reports looking at many papers on a particular topic. It was named after Archie Cochrane, a doctor who was captured by the Germans in World War 2 and somehow managed to organize his fellow prisoners in a trial to test the effects of different diets!
6. Ask who is missing - look at how the data was collected and whether it was on a representative sample of people or not.
7. Demand transparency when the computer says no - don't blindly trust algorithms, as they're only as good as the data they're fed. One interesting thing I've read but had forgotten - the average body temperature is more like 98.1 degrees than the commonly-known 98.6 degrees, partially because the doctor did his measurements in Celsius and converted it to Fahrenheit, and partially because it sounds like the doctor took people's temperatures in their armpits and also maybe his thermometers were miscalibrated? But he had a ton of data and so people just trusted it for a long time...
8. Don't take statistical bedrock for granted - this chapter is basically about how important government statistics are, and how important it is that we can trust them. He talks about how under Margaret Thatcher the UK's definition of "unemployment rate" changed more than 30 times in a decade, generally to make the rate look lower. This of course damaged the reputation of the UK statistics office for quite a while.
9. Remember that misinformation can be beautiful, too - it's easy to create pretty graphics that are misleading.
10. Keep an open mind - it's natural to discount facts that disagree with your preconceptions, so, um, try not to do that.
Maybe it's because I've read a lot of this stuff before in different forms. Maybe a year and a half in a pandemic I'm tired of reading through virus statistics in the news and trying to make sense of it. Maybe I'm depressed about the state of the world!
...anyway, there were some interesting bits. Harford starts off by talking about the book "How to Lie with Statistics" by Darrell Huff, which was published in 1954 and became exceptionally famous for a book about statistics. (I had a copy growing up!) Reading it makes you more skeptical about numbers you see in the press.
But! Another thing that happened in 1954 was that one of the first studies that showed smoking cigarettes caused lung cancer came out. And we needed statistics to figure that out. And guess who testfied before Congress 11 years later on behalf of the cigarette companies, arguing that the studies were wrong and cigarettes were fine? The same Darrell Huff.
This is a powerful story! Harford's argument is that while we should be careful, we shouldn't be too cynical about statistics, because they can show us things that are almost impossible to see otherwise. The rest of the book is a list of rules to follow when you're looking at statistics to give yourself a better chance of understanding them:
1. Search your feelings - try to avoid emotional responses to data because it can cloud your judgment. (if you see a graph that you agree with, be extra careful before sharing it!)
2. Ponder your personal experience - if your personal experience contradicts some statistics you're seeing, maybe you should take a closer look at the statistics. Harford gives the example of measuring the average number of riders on a London Underground train throughout the day, which was pretty low. But his personal experience is that every time he takes a train to work it's extremely crowded. But these can both be true if most people are taking the train at the same time, and perhaps a better metric is the average number of other people on a train when a person gets on, or something like that. But conversely, lung cancer is rare enough that you can't trust your own intuition; you have to look at the statistics. I'm not real clear on how you're supposed to differentiate between these cases...
3. Avoid premature enumeration - be careful about what's actually being counted (for example, there were ~40,000 gun deaths in the US in 2017, but ~60% of these are suicides, which can definitely frame how you think about the issue differently)
4. Step back and enjoy the view - getting perspective is important, either by looking at different times, or asking "is that a big number?". Harford also points out that bad news tends to overpower good news in the media, and one reason for that is that bad news tends to happen suddenly, while good news happens slowly over time.
5. Get the backstory - it's common to see outlier successes without seeing all the common failures. (~40% of Kickstarter projects don't meet their goal, but if you hear about one in the media it's one that's been a huge success) Harford also talks about publication bias, and the fact that there's also a bias to publish novel and surprising discoveries but not replication attempts that might debunk previous "discoveries". One interesting point is that the "backfire effect" (where fact-checking can make people more likely to believe a false claim) is actually unusual and fact-checking does help in most cases. This chapter also talks about the Cochrane Collaboration, which is a group that looks at published medical papers and creates research reports looking at many papers on a particular topic. It was named after Archie Cochrane, a doctor who was captured by the Germans in World War 2 and somehow managed to organize his fellow prisoners in a trial to test the effects of different diets!
6. Ask who is missing - look at how the data was collected and whether it was on a representative sample of people or not.
7. Demand transparency when the computer says no - don't blindly trust algorithms, as they're only as good as the data they're fed. One interesting thing I've read but had forgotten - the average body temperature is more like 98.1 degrees than the commonly-known 98.6 degrees, partially because the doctor did his measurements in Celsius and converted it to Fahrenheit, and partially because it sounds like the doctor took people's temperatures in their armpits and also maybe his thermometers were miscalibrated? But he had a ton of data and so people just trusted it for a long time...
8. Don't take statistical bedrock for granted - this chapter is basically about how important government statistics are, and how important it is that we can trust them. He talks about how under Margaret Thatcher the UK's definition of "unemployment rate" changed more than 30 times in a decade, generally to make the rate look lower. This of course damaged the reputation of the UK statistics office for quite a while.
9. Remember that misinformation can be beautiful, too - it's easy to create pretty graphics that are misleading.
10. Keep an open mind - it's natural to discount facts that disagree with your preconceptions, so, um, try not to do that.
October 24, 2021
Tim Harford is an economist, best known to the general public as presenter and participant in television and radio programmes such as BBC's "Trust Me, I'm an Economist". Indeed, "How to Make the World Add Up" often references the BBC 4 "More or Less", a programme about the accuracy of numbers and statistics in the public domain. The book takes a similar approach, in that, without in any way undermining the usefulness of statistics in understanding the world around us, Harford approaches the subject with a healthy scepticism. He sets out ten rules which can help the reader question public statements based on statistics and arrive at realistic conclusions unbiased by personal prejudice or media and political spin.
April 21, 2022
I enjoyed this book, but I’m not really up for writing a full review. So let’s see what I can piece together from my sparse notes. I can’t find a first-rate review, either here or online, so how to convince you the book is worth reading? Understanding the basics of interpreting statistical claims is pretty straightforward: be skeptical of grand claims. Slow down and think things through. Try not to fall in love with a claim that you find attractive just because it’s comfortable. Be curious: look deeper and ask questions. Harford thinks cultivating curiosity is the best tool, the Golden Rule. If this awakens your sense of wonder, so much the better!
Harford writes well and is an experienced and fluent pop-science writer. He did irritate me early on by berating skeptics of some of the more extreme claims about Climate Change as “Climate Deniers,” violating his own rules. Hey, nobody’s perfect. I think you are likely to get more out of the book if you have some prior experience with using statistics, but any curious reader is likely to find themselves enjoying his stories. I certainly did, and I have training and experience in this field, long long ago. Who knew that Florence Nightingale was a skillful and effective statistician?
Recommended for numerate readers.
Harford writes well and is an experienced and fluent pop-science writer. He did irritate me early on by berating skeptics of some of the more extreme claims about Climate Change as “Climate Deniers,” violating his own rules. Hey, nobody’s perfect. I think you are likely to get more out of the book if you have some prior experience with using statistics, but any curious reader is likely to find themselves enjoying his stories. I certainly did, and I have training and experience in this field, long long ago. Who knew that Florence Nightingale was a skillful and effective statistician?
Recommended for numerate readers.
June 4, 2021
Very well written guide to help average people read, understand, and question statistics they may run across. Totally non-technical, with gentle understated humor, there are no formulas and nothing to frighten a math-phobe. He also makes it clear he’s not out to debunk statistics. His position is that statistics are a great tool to understand reality, but they’re frequently imperfect, so it’s good for us non-technical people to understand how they can go wrong. He ends up with a nice chapter about how as consumers of statistics our best strength is our curiosity. (In my reviews I often carp that the last chapter or two of non fiction books are substandard - not with this book, it’s uniformly enlightening and enjoyable to read.)
June 14, 2021
The key message :
To look at any kind of data with a clear mind and a focus on the facts, remember some important rules. These include watching your emotional reactions to information – whether visual or verbal –and being willing to update your opinions in the face of new evidence. You should also look at the big picture of a statistic, making sure to examine the overarching context and identify potential distortions, exclusions, or oversights. The ultimate goal tying all of these together is to always be curious – look deeply for the facts and keep asking questions.
Actionable advice:
Memorize a few “landmark numbers.”
Entrepreneur Andrew Elliott advocates keeping a short list of “landmark numbers” in your head so it’s easier for you to understand the relative significance of other numbers. Here are a few examples: The population of the United States is 325 million; the UK’s is 65 million. The drive from Boston to Seattle is 3,000 miles. And the average novel is 100,000 words long. Once you’ve got those in your head, you can use them to make comparisons – for instance, a 10,000-word report might seem long, but it’s ten times shorter than the average novel.
---------دیگه به الگوریتم کلان داده ها هم نمیشه اعتماد کرد
Maintain a healthy skepticism of algorithms and big data.
Upon its release in 2009, Google Flu Trends was touted as a revolutionary tool in tracking the spread of seasonal influenza. By counting searches for “flu symptoms” and “pharmacies near me,” Google could accurately estimate new daily flu cases faster than the CDC.
Google Flu Trends was in many ways the herald of a new age: that of “big data” and algorithms. Big data refers to the information we produce when we surf the web, pay with credit cards, or use mobile phones. Algorithms are computer programs often used to find patterns in datasets.
Google Flu Trends used big data and algorithms to –it seemed – produce good data on flu trends. Yet, just four years after the project was announced, it completely collapsed. Why?
Google Flu Trends crashed and burned when, one winter, it suggested there was a severe outbreak when there wasn’t one. At one point, it estimated that the spread of flu was two times worse than suggested by the official CDC data.
So what was the problem? Mainly it was that Google didn’t, in fact, know what the connection was between search terms and the spread of flu. The algorithm was searching for patterns in the data, but it found connections involving things unrelated to flu, such as “high school basketball.” As a result, the algorithm became less of a flu detector and more of a general-purpose winter detector. That meant it was unable to detect a summer outbreak of flu that occurred in 2009.
Of course, there are sometimes cases in which it is worth trusting algorithms over human-produced estimates. For instance, there’s a wealth of evidence to suggest that human judges are neither wholly objective nor consistent when making criminal sentences. Algorithms are much better at producing fair sentences by comparing cases to similar ones in the past.
Sometimes algorithms will produce accurate, quality results, and sometimes they won’t. Consequently, we’ll need to judge each algorithm on a case-by-case basis and not take its accuracy as a given.
Doing this can be difficult because many companies don’t want to reveal the secrets behind their money-making engines. But if everyone is allowed to peer under the hood of an algorithm, we’ll be much more likely to understand how they’re making the decisions they are –and how they can improve.
(blinkist.com)
To look at any kind of data with a clear mind and a focus on the facts, remember some important rules. These include watching your emotional reactions to information – whether visual or verbal –and being willing to update your opinions in the face of new evidence. You should also look at the big picture of a statistic, making sure to examine the overarching context and identify potential distortions, exclusions, or oversights. The ultimate goal tying all of these together is to always be curious – look deeply for the facts and keep asking questions.
Actionable advice:
Memorize a few “landmark numbers.”
Entrepreneur Andrew Elliott advocates keeping a short list of “landmark numbers” in your head so it’s easier for you to understand the relative significance of other numbers. Here are a few examples: The population of the United States is 325 million; the UK’s is 65 million. The drive from Boston to Seattle is 3,000 miles. And the average novel is 100,000 words long. Once you’ve got those in your head, you can use them to make comparisons – for instance, a 10,000-word report might seem long, but it’s ten times shorter than the average novel.
---------دیگه به الگوریتم کلان داده ها هم نمیشه اعتماد کرد
Maintain a healthy skepticism of algorithms and big data.
Upon its release in 2009, Google Flu Trends was touted as a revolutionary tool in tracking the spread of seasonal influenza. By counting searches for “flu symptoms” and “pharmacies near me,” Google could accurately estimate new daily flu cases faster than the CDC.
Google Flu Trends was in many ways the herald of a new age: that of “big data” and algorithms. Big data refers to the information we produce when we surf the web, pay with credit cards, or use mobile phones. Algorithms are computer programs often used to find patterns in datasets.
Google Flu Trends used big data and algorithms to –it seemed – produce good data on flu trends. Yet, just four years after the project was announced, it completely collapsed. Why?
Google Flu Trends crashed and burned when, one winter, it suggested there was a severe outbreak when there wasn’t one. At one point, it estimated that the spread of flu was two times worse than suggested by the official CDC data.
So what was the problem? Mainly it was that Google didn’t, in fact, know what the connection was between search terms and the spread of flu. The algorithm was searching for patterns in the data, but it found connections involving things unrelated to flu, such as “high school basketball.” As a result, the algorithm became less of a flu detector and more of a general-purpose winter detector. That meant it was unable to detect a summer outbreak of flu that occurred in 2009.
Of course, there are sometimes cases in which it is worth trusting algorithms over human-produced estimates. For instance, there’s a wealth of evidence to suggest that human judges are neither wholly objective nor consistent when making criminal sentences. Algorithms are much better at producing fair sentences by comparing cases to similar ones in the past.
Sometimes algorithms will produce accurate, quality results, and sometimes they won’t. Consequently, we’ll need to judge each algorithm on a case-by-case basis and not take its accuracy as a given.
Doing this can be difficult because many companies don’t want to reveal the secrets behind their money-making engines. But if everyone is allowed to peer under the hood of an algorithm, we’ll be much more likely to understand how they’re making the decisions they are –and how they can improve.
(blinkist.com)
July 12, 2022
A wonderful book with a clear passion for numbers and more specifically, communicating complex ideas correctly with numbers. The author walks us through 10 rules that we should keep in mind when confronted with statistics or graphs, but I most enjoyed the 11th chapter where he honestly admits that '10 rules' is a bit silly and that all rules can be summarize as 'be curious.' That last chapter is a heartfelt plea to be curious about the world and offers a number of insights from research that demonstrate that this open and curious mindset could save our polarising world. In these sometimes dark times, it was thrilling to read such a hopeful and passionate message. The ten chapters before that are also a thoroughly enjoyable read as the author clearly knows how to weave a chapter together (combining historical anecdotes, personal observations and research examples into a fluent read).
February 12, 2022
4.5/5 This was quite similar to books like Thinking, Fast and Slow which is ofcourse the best book on the subject. Ten chapters/rules and in each chapter the author illustrates how we can be misled by an headlines and statistics.
In particular, I loved learning how artificial-intelligence algorithms that seemed to be racially biased in their outcomes were actually so because - they were racially neutral. Different people behave differently according to their culture and upbringing, and algorithms actually missed it because they treated people the same. The irony of it !
Will be reading all by the author I think !
In particular, I loved learning how artificial-intelligence algorithms that seemed to be racially biased in their outcomes were actually so because - they were racially neutral. Different people behave differently according to their culture and upbringing, and algorithms actually missed it because they treated people the same. The irony of it !
Will be reading all by the author I think !
January 27, 2023
The title sums it up: Harford provides ten easy rules to make sense of statistics or any numerical information for that matter. Well written with well known stories and references to related literature.
“Yes, it’s easy to lie with statistics—but it’s even easier to lie without them.”
“I worry about a world in which many people will believe anything, but I worry far more about one in which people believe nothing beyond their own preconceptions.”
“Yes, it’s easy to lie with statistics—but it’s even easier to lie without them.”
“I worry about a world in which many people will believe anything, but I worry far more about one in which people believe nothing beyond their own preconceptions.”
May 20, 2023
Ok, a bit too simple.
August 11, 2021
Tim Harford nimmt uns mit in die Welt der Daten und Statistiken und wie sie manchmal unsere Urteilsfähigkeit trüben. Was ist es was wir falsch machen, wenn wir Daten analysieren, kann man Big Data 100% vertrauen?
Anhand von vielen Beispielen und den Wirtschaftspsychologischen Theorien von Daniel Kahnemann, kriegt man wieder einen schönen Einblick wie kognitive Verzerrung funktioniert.
Anhand von vielen Beispielen und den Wirtschaftspsychologischen Theorien von Daniel Kahnemann, kriegt man wieder einen schönen Einblick wie kognitive Verzerrung funktioniert.
March 20, 2023
As a data person and self professed skeptic I loved this! Numbers are used and misused with such growing frequency to persuade us about policies, how we spend our money, and how we live our lives that I think everyone should read this book to help filter through these stats more effectively.
August 23, 2022
THIS IS NOT A BOOK ABOUT THE MATH OF STATISTICS. NO FORMULAS NEEDED, This is a very good book that many types of readers will enjoy. It is really more about the psychology of statistics and what questions to ask when you see data to help you decide if the data really reflect what is claimed.
Some of it was a bit shocking but on second thought not really surprising. For example, there was a study (ONE study) that seemed to show that people have precognition published in the Journal of Personality and Social Psychology. It was so unexpected that several other researchers tried to replicate it, but the studies all failed. None of the researchers could get THEIR results published, though, because the journal said they "did not publish replications". Such experiences discourage researchers from trying to confirm or disprove questionable results, especially younger researchers without tenure and reputation. So the journal seems overly disposed to publish a surprising result but overly reluctant to publish something that shows it was wrong.
Some of it was a bit shocking but on second thought not really surprising. For example, there was a study (ONE study) that seemed to show that people have precognition published in the Journal of Personality and Social Psychology. It was so unexpected that several other researchers tried to replicate it, but the studies all failed. None of the researchers could get THEIR results published, though, because the journal said they "did not publish replications". Such experiences discourage researchers from trying to confirm or disprove questionable results, especially younger researchers without tenure and reputation. So the journal seems overly disposed to publish a surprising result but overly reluctant to publish something that shows it was wrong.
February 28, 2021
There are a fair number of books on statistics that are regularly published: a very good one last year was ‘Calling Bullshit’ by Carl T. Bergstrom & Jevin West. Like ‘Calling Bullshit’ Data Detectives is at times presenting the reader with statistical background on various unrelated topics to explore the use of statistics, but other times Harford is engaging more in cherry picking data and interpreting it to fit his own worldview. His take on income inequality is one such example where he lazily chooses to look at one statistical factor (Ginny coefficient) and call it a day. Considering his employment from the conservative Financial Times it doesn’t surprised that he would hold this view, but it is surprising that he wouldn't look much deeper into this topic.
September 26, 2020
An excellent contribution to the debate from Tim Harford. Although he might have had to twist things a little to arrive at ten rules, his style, stories and approach make for a great read. Quite a bit wasn't exactly new to me, but even when I knew something about a topic or story, he menaged to add to it. Only thing missing was Twyman's Law of statistics and data analysis, which Tony Twyman himself introduced me to over forty years ago: "if it looks interesting, it's probably wrong…" And one main reason why I dislike multivariate analysis and neural networks so much: you can't see what's going on.
May 2, 2023
Great points on how to read any statistics, analysis and get insight. Be curious, look whats missing, remove emotions and what/who is the source, and many more…
June 10, 2023
A wonderfully accessible book that teaches cognitive hygiene - by which I mean: how to think clearly and go beyond the surface of things. The "ten easy rules" themselves are rather vague and sound more like self-help mantras. But the real value in the book lies in the examples and case studies, which show how easily misled we are by our own wishes and inclinations, by fancy graphics, by overestimating our own knowledge and underestimating our own biases, and by failing to understand what exactly is being measured and represented.
This is a book about thinking and interpreting data, not about the mathematics of statistics. As a matter of fact, I think there is not a single equation. What I like best about the book, though, is that it is not defaitist or gloomy, along the lines of "statistics are imperfect, therefore let's ignore them and just believe what we choose to believe". Instead, it offers a positive message: "If you make the effort to dig a little bit under the surface, you may find very useful information in statistics, which may help you with the decision-making that you need to do".
The author sounds like the type of person you'd like to have on speed-dial every time you read a newspaper article or demographic survey that surprises you. Or perhaps even more importantly: whenever you read something that DOESN'T surprise you, and where you may be falling victim to confirmation bias.
This is a book about thinking and interpreting data, not about the mathematics of statistics. As a matter of fact, I think there is not a single equation. What I like best about the book, though, is that it is not defaitist or gloomy, along the lines of "statistics are imperfect, therefore let's ignore them and just believe what we choose to believe". Instead, it offers a positive message: "If you make the effort to dig a little bit under the surface, you may find very useful information in statistics, which may help you with the decision-making that you need to do".
The author sounds like the type of person you'd like to have on speed-dial every time you read a newspaper article or demographic survey that surprises you. Or perhaps even more importantly: whenever you read something that DOESN'T surprise you, and where you may be falling victim to confirmation bias.
Displaying 1 - 30 of 726 reviews