This excellent book is something very unusual.
First, it's about numbers but manages to be both extremely easy to read and very entertaining.
Secondly, although it is so accessible that a ten-year old of average intelligence should be able to understand everything in this book, the points it makes are so universal in application that even someone with much greater mathematical knowledge - and I write this as a graduate with two degrees in a discipline which requires statistical understanding - can find it full of useful reminders and even the odd valuable idea you might not have thought of or heard of.
The book is about how numbers can be manipulated, by accident or design, to trick people into making false conclusions, and how to spot when you are being fed misleading numbers. In this day and age the ability to spot bad statistics is extremely important to everyone and can literally be a life-saver.
I was very surprised indeed to see that a previous reviewer had described this book as "not for everyone." I could not disagree more strongly.
If every voter read this book, fewer bad politicians would be elected on the basis of dishonest campaign statistics, if every consumer read it, fewer bad products would be sold on the basis of dishonest advertising statistics, and if every journalist read it there might be less harm done by scare stories based on bad statistics.
Despite the fact that this book was written many years ago, every single word in it is still very relevant today.
However, anyone with a serious interest in the subject who wants an update on some of the more recent examples of how statistics are misused should still start by reading "How to Lie with Statistics" and then follow up with the equally good "Damn Lies and Statistics" by Joel Best, which is more current and nearly as accessible. The two books complement each other very well.
on 10 February 2005
I first read this book when I was about 12, and re-read it now that I'm in my 20s, and am amazed by how good it is. It's got the complexity of a textbook, but the writer has no pretensions and has managed to get the information across in a way so simple a child can easily read the book and understand some of his lessons and examples.
There are plenty of lessons about how we should interpret the numbers we come across every day in adverts and (potentialy biased) news reports and there is nobody living in the developed world who can't benefit from the enlightenment that this brings.
The only disappointing aspect of this book is that it's so short, an accomplished reader with some knowledge of statistics could get through the book in a single (if lengthy) sitting.
on 17 July 2002
While anyone who has dealt with statistics in a professional capacity is probably familiar with the contents already, it is still a handy little reference. And for anyone in an introductory course of study or who is simply concerned enough to wonder about the truth of what they read, this is absolutely invaluable.
It is not a long book, and some of the examples are dated (physicians recommending brands of tobacco, for instance), but the meat of the book is both accurate and extremely readable. It covers the ways that statistics can be made to show pretty much anything, both through deliberate manipulation and through simple sloppiness. The main chapters cover issues such as inadequate and biased samples, how to provide subtly and not-so-subtly misleading (though technically accurate) visual charts and representations, how to manipulate perception by eliminating inconvenient precision and adding spurious precision, how to manipulate perception by supplying numbers without context or by simply leaving inconvenient facts out, and how to confuse people thoroughly about correlation vs. cause-and-effect. The final chapter provides a nice summary: the questions you absolutely must ask about any figure you are presented with, in order to judge its worth.
As the author himself says, it may read something like a graduate text on dishonesty, but one can assume that people who deliberately wish to mislead have figured out how already; this is to educate the honest person who wishes to be alert. It is frequently used as a text in undergraduate statistics courses, for good reason.
on 6 November 2003
I first bought this book in 1977 when I was doing an Open University course. It is still as useful now as it was then. Anyone concerned at the spin and lies that gush from our friends in the government should possess this book. An approachable and essential guide to bulls**t detection.
on 11 November 2003
Don't be put off by the catchy title. This engaging short book will help you to see your way through graphs and understand how to manipulate numbers appropriately - and it will do it easily, without losing you in mathematical complications. I loved this book when I first read it at school, still loved it through my maths degree, and now that I've forgotten all my maths I go back to it occasionally to refresh myself.
on 18 December 2008
There is still much to lament in our ruling classes, but thank goodness we are no longer led by the likes of Lord Randolph, whose epiphany with regard to the decimal point is quoted above. However low your opinion of contemporary politicians, Gordon Brown is unlikely to mistake ".34 per cent" for "34 per cent" (although our financial ruling class would probably reward themselves handsomely for being out by only a factor of a hundred). Before we congratulate ourselves too readily for our mathematical sophistication, we should reflect upon the salutary fact that Darrell Huff's classic text is as necessary today as it was when it was first published over half a century ago. It is remarkable that, despite certain figures showing their age (we might be heading back to an average wage of £1,400 but we're not quite there yet), there is nothing dated about his style. A maths book on statistics from the fifties? If this seems as appetizing as a cold spam butty during a power cut, you're in for a surprise. That decade was not entirely in black and white.
If you spotted the fast one I pulled in the first paragraph, you're either one of "the crooks" who already know these tricks or else are an honest soul who has learned them "in self-defence". Hence the title of this fantastic little book: knowing how a burglar thinks helps secure your house. Most of the time, I would pass over the phrase "average wage" without a second glance. We all know what an average is, don't we? Distant maths lessons are just that for most of us, and even if I'd dredged up the question - what kind of average? - would I have been bothered to ask it? Complacency translates into vulnerability.
"When you are told that something is an average you still don't know very much about it unless you can find out which of the common kinds of average it is - mean, median, or mode." Without a clear understanding of these different kinds of average, you have to hope it doesn't really matter which one is being used, but this is only the case "when you deal with data... that have the grace to fall close to what is called the normal distribution." Otherwise, it makes a big difference, so much so that, "as usually is true with income figures, an unqualified 'average' is virtually meaningless."
Advertisers, of course, are among the most culpable and capable when it comes to lying with statistics (although at least their motives are plain). It is typical of Huff's sense of mischief that, alongside the calculations, he presents us with an ethical dilemma of enormous proportions: should we feel sorry for advertisers who are themselves victims of statistical skulduggery? For example, a magazine publisher is happy to state the median age of its readership, while leaving the kind of average for incomes "carefully unspecified". "Could it be that the mean was used instead because it is bigger, thus seeming to dangle a richer readership before advertisers?"
This is a short book, made even shorter by pictures of cows and charts that take up half a page. (How the innocent-looking graph can be manipulated by adding "schmaltz" is another example of Huff's style: a simple unpicking of the familiar to demonstrate an important point.) It is also unreasonably funny in parts. I don't recall maths, let alone statistics, ever being this entertaining at school. And yet the intellectual content is not compromised. Huff's message is a serious one and perhaps more important now, since our propensity for attaching numbers to almost anything shows no sign of diminishing. It ought to be common knowledge that samples can be "biased by the method of selection", that "well-biased samples can be employed to produce almost any result anyone may wish", that it can be difficult to obtain "a representative sample... one from which every source of bias has been removed", that people who answer survey questions have "a desire to give a pleasing answer", that strange results "crop up when figures are based on what people say".
Most of us can understand these ideas when they are explained by someone like Huff (although it might help not to be an aristocrat). If we're honest, out in the wild without a guide, we're not so sure. Have you ever been scared by "accident statistics"? Would the fact that more people "were killed by aeroplanes last year than in 1910" give you pause for thought? Are modern planes really more dangerous? "Nonsense. There are hundreds of times more people flying now, that's all."
"It is sometimes a substantial service simply to point out that a subject in controversy is not as open-and-shut as it has been made to seem." Or, as Goldacre's slogan has it: "I think you'll find it's a bit more complicated than that..." Both belong to that noble tradition of satire with a serious message, and it is a tribute to Huff's writing style that he can end with a quote from Mark Twain that fits perfectly: "There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact."
on 1 May 2002
An excellent resume of the ways that statistics can be used to mislead without, technically, lying. I read my copy every couple of years or so just to refresh my cynicsm. Everyone with an interest in advertisng, politics, PR should read it.
I love this book. Short, sweet, to the point.
In our modern world of spin and advertising this book is a valuable antidote. In tests 8 out of 10 readers said they preferred this book...After reading it you'll know to ask "preferred it to what?"
This book is a valuable and valid aid to those who prefer truth to fiction. Children just getting to grips with statistics will enjoy it, and adults will remind themselves of much basic good sense.
on 6 January 2015
I could talk about how this book is a primer in how to think clearly, with Huff exposing errors in reasoning ranging from the post hoc fallacy to fallacious appeals to authority. I could also talk about how the author uses real life examples at every given opportunity, supplemented by neat diagrams and pictures. I could also talk about how this book is only 140 pages, making it ideal for light reading.
Yes, I could talk about all of the above things in great detail, but I want to do something different, albeit slightly longer- I want to give you a chapter by chapter summary of this great work which is still relevant today.
Chapter One: The sample with the built in bias.
The Gist of this chapter is the self-evident truth that people have a tendency to exaggerate. If you rely on what people say that they do as opposed to what they actually do as a basis for your study, your study will be biased from the outset.
Chapter Two: The well-chosen average
When someone tells you that figure X is an "average", ask them "what average are you talking about?"
Is it the mean, median, mode or some other measure?
Chapter Three: The little figures that are not there
If you repeat your test enough times, you will eventually get the result you want through sheer chance alone. You then have the luxury of only publicising the cases where you achieved your desired results.
Chapter Four: Much ado about practically nothing
The IQ as a product of a sampling method has a statistical error. It could actually be the case that two individuals who are ranked differently in terms of IQ are no different in intellect, after accounting for these errors.
Chapter Five: The Gee-Whiz graph
If you want to make your argument more "convincing", just play around with your graphs.
For instance, if you want to exaggerate a trend line- just change the scale on the Y axis.
Chapter Six: The One-Dimensional picture
This is quite similar to chapter five but Huff talks about how we can use images instead of graphs in order to overstate reality.
Chapter Seven: The Semi attached figure
You can't prove that your breakthrough medicine cures colds, but you can show that it kills 30,000 germs in 11 seconds whilst ignoring that the effects inside of a test tube are not necessarily the same as what happens within the human throat.
Chapter Eight: Post Hoc rides again
Correlation does not equal causation.
Chapter Nine: How to Statisticulate
Huff coins this term which is a combination of the words "statistical" and "manipulation".
Huff shows how those within a position of power can use statistics to manipulate the public. Huff asks the question: "When has a union ever employed a statistician so incompetent that he made labor's case out weaker than it was?"
In short, as long as the "errors" remain one sided it is not easy to attribute them to accident.
Chapter Ten: How to talk back to a statistic
Much like a retired magician who exposes secrets about the "magical" industry, Huff arms the reader with the tools needed to see through statistical manipulation.
on 22 April 2004
The worst part of this book is the cover - it nearly stopped myundergraduate son from reading it! I had remembered it from my ownundergrad days. I bought it to re-read on holiday - yes, it IS that easyto read - and my son sneered as he walked by.
So I engaged him in debate, using one of the examples from the book. Badmistake. I didn't see the book again for the next two days; it's slim butvery thought provoking so he'd read a bit and come and argue with me aboutit. He nearly damn stole the book at the end of our break.
Everyone should read this book in young adulthood.