The Lowdown: This How to Lie with Statistics Summary is based on the popular book from Darrell Huff, which demonstrates how all numbers can be manipulated, even when they are presented as objective.
Anytime a person wants to make a strong case for why an idea is true, they often rely on statistics. Because numbers themselves can’t lie, we often times take the stats and accept them as true and accurate, because often times they are put together by experts.
However, as with anything in life, it is never that simple. Very rarely are stats built within an unbiased vacuum, and due to this, they will always never be 100% accurate.
- Audible Audiobook
- Darrell Huff (Author) – Bryan DePuy (Narrator)
- English (Publication Language)
- 02/24/2016 (Publication Date) – Audiobooks.com Publishing (Publisher)
The Three Main Lessons you’ll learn from How to Lie with Statistics include:
- The way we gather information for Stats is inherently wrong
- How Charts can be Manipulated
- Difference between Percentages and Numbers
Lesson One: The way we gather information for Stats is inherently wrong
Often times we hear of studies and the information they provide, and accept the values for what they are. However, we never peel the layer back to see how they gathered the information.
For example, how many people did they ask? Did they call these people, and if so, what time of day were they called? Did they send an e-mail and have people take a survey, or did they do a social media poll to find the information? Were they mostly women or men? What ages were the people?
There is very little transparency within studies, and due to this, the information which is gathered doesn’t tell the full story. There are typically zero controls within the group, meaning the information which is gathered cannot possibly be completely unbiased. You could conduct a study five different times, and because the people are completely random, could have five different results. There must be certain controls put into place in order for a study to have a meaningful impact, otherwise, the information which is gathered provides no value.
Lesson Two: How Charts can be Manipulated
If you read a story on-line, or watch a news show, they often times will use a graph to demonstrate the results they are showing. However, most graphs can be manipulated to accentuate the information they really want to show.
For example, if you want to show the difference in wages between men and women, and the difference is $10,000 a year (say, in a hypothetical sense, Women make $90,000 a year and Men make $100,000 a year), you could show a bar graph. The bar graph can start at $0, and with that, you can see there is a difference but the gap doesn’t look like much. However, if you start the graph at $75,000, and use the same size of graph, then the difference will look a lot bigger.
You can also use other graphics to showcase this difference. Rather than a bar chart, you can use a bag of money to represent the difference in value. Not only would the $60,000 be taller, but since it is taller, in order to look correct, it would also have to be wider. These ways of manipulating graphs and graphics lead to a bias, and help shape a message which may not b completely accurate.
Lesson Three: Difference between Percentages and Numbers
If we use the same example as above, it can be stated in a few different ways.
Women make 10% less than Men on an average basis.
Over their working careers, Women will make over $450,000 less than Men.
Women make an average of just under $1,000 a month less than Men.
There are different ways to convey the results, and depending upon who is delivering the results will typically determine how inflammatory the information really is.
Not only this, but hopefully, you have started to ask some questions concerning this data. For example, how many women and men did they survey? What fields do the men and women work within which they received this information from; are they comparing the same industries? Are the men and women they are asking the same age?
Without knowing all the variables which go into each study, then reading the data (whether it is presented by percentages or by numbers) can be very misleading.
My Personal Takeaway
When it comes to surveys and polls, it is clear that I do not ask enough questions. By typically only showing the results of even one question, it makes you ponder how many questions were the people asked? What are the controls which are being used to ensure the information which is presented is fair and accurate?
Without having this information, then it is very difficult to take any survey at face value.
Put into Action
Pay more attention to the information being presented, and become a critical thinker. Don’t take the information as 100% complete truth until you ask the right questions. A medication might show that it can reduce your chances to contract a common cold, but will it increase your chances for heart disease or another ailment? Ask all the necessary questions before enacting upon the information given.
You should consider buying this book if…
You want to learn more about how statistics are used to manipulate the information you see on a day to day basis, and become more aware of how to start challenging the information which is given to you.
a Swedish university student, marketing professional, and life-long learner. Here at BookSummaryClub I summarize my favorite non-fiction books into easily digested posts. Hope you like what you’re reading!
🤙 Your Next Step… 🤙
Head across to one of the following pages for more goodies
🍕 Read our Blinkist review and become a member of Blinkist. Read or listen to 3000+ full version quality summaries!
🍕 Read our list of the best business books of all time
🍕 Read some more of our book summaries
🍕 See our top book summary apps