How to Lie With Statistics by Darrell Huff
A great reminder of the basic principles of statistics and how they are often violated in the data we use and see in work, the news, and our personal lives.
Buy this book on Amazon (Highly recommend)
Learn and be cautious with statistics
“The secret language of statistics, so appealing in a fact-minded culture, is employed to sensationalize, inflate, confuse, and oversimplify.”
In our “science-backed” society, we are inundated with statistics. While some of these facts and figures may be valid, often they are incorrect or misleading ways to convince us of the agenda of some person or organization. So instead of believing every statistic you see or rejecting numbers entirely, learn the language of statistics and the common ways in which it is misused and abused so that you can recognize when you might be at risk for being duped.
On the difficulty of sampling
“To be worth much, a report based on sampling must use a representative sample, which is one from which every source of bias has been removed.”
“The operation of a poll comes down in the end to a running battle against sources of bias, and this battle is conducted all the time by all the reputable polling organizations. What the reader of the reports must remember is that the battle is never won. No conclusion that “sixty-seven percent of the American people are against” something or other should be read without the lingering question, Sixty-seven percent of which American people?”
For any statistic to be worth its salt, it should come from a representative sample of the group about which it is describing. However, getting a truly representative sample is almost impossible for time, money, and other practical reasons. Therefore, when you see statistics that claim to accurately describe a particular group, remember that depending on the sampling method used, it’s likely that the statistic is describing some subset of the group in question, rather than the group itself. For example, for every poll that claims to describe the beliefs of an entire nation, it’s likely that the statistic represents only some part of that nation based on the biases coming from the method of sampling. To mitigate the risk of believing these statistics accurately represent the entire nation, always ask yourself, “Which part of that nation?” when you see these types of statistics.
What kind of average?
“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.”
When a statistic cites the “average,” remember that it might be the mean, median, or mode, which can be radically different. Often, a person who uses the unspecified “average” will use the number that most clearly proves his or her point while leaving out any numbers that might bring that point into question. So if you see an average that does not specify whether it is the mean, median, or mode, do your homework and figure out what it is.
What is the test really measuring?
“…whatever an intelligence test measures it is not quite the same thing as we usually mean by intelligence. It neglects such important things as leadership and creative imagination. It takes no account of social judgment or musical or artistic or other aptitudes, to say nothing of such personality matters as diligence and emotional balance.”
If you’re going to use a test to measure somebody or something, think deeply about what that test does not measure. For example, an intelligence test does not measure many of the qualities that we associate with real-world intelligence, such as leadership, creative imagination, diligence, and emotional balance. That doesn’t mean that an intelligence test is not useful, but it does mean that it does have significant limitations that matter if you want to accurately evaluate the “intelligence” of a person, whether that is for admission to a university or a job.
Statistics are biased towards one side
“As long as the errors remain one-sided, it is not easy to attribute them to bungling or accident.”
Nobody uses statistics to disprove themselves. That is, statistics are used to serve a cause or agenda that someone believe ins, rather than disprove that cause or agenda. So when people err in their use of statistics or make intentional mistakes, the errors will always be in one direction (the direction of their beliefs and agenda). So instead of getting an unbiased look at the world, we’re getting a biased look of the world that other people want us to see. This is a dangerous world to live in if you don’t recognize this reality about how people use data.
Does it make sense?
“‘Does it make sense?’ will often cut a statistic down to size when the whole rigmarole is based on an unproved assumption.”
If you hear a statistic that seems off, or if someone arrives at a conclusion that seems questionable, ask yourself, “Does it make sense?” Instead of accepting what people have to say, even if those people are really smart, think about what it is that they are saying and why that might be. In the end, they might be right, but it’s worth stopping and asking the question with a genuine sense of curiosity.
If you found these notes helpful...
You might also enjoy these books...
- How to Stop Worrying and Start Living by Dale Carnegie
- Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets by Nassim Taleb
- Thinking in Bets: Making Smarter Decisions When You Don’t Have All the Facts by Annie Duke
- Think Again: The Power of Knowing What You Don’t Know by Adam Grant
- Start with Why: How Great Leaders Inspire Everyone to Take Action by Simon Sinek
- Radical Candor: Be a Kick-Ass Boss Without Losing Your Humanity By Kim Scott
- The Ride of a Lifetime by Robert Iger
- Measure What Matters: How Google, Bono, and the Gates Foundation Rock the World with OKRs by John Doerr
- The Man Who Solved the Market: How Jim Simons Launched the Quant Revolution
- Steve Jobs by Walter Isaacson