How to Lie with Statistics

How to Lie with Statistics

by Darrell Huff
3.84 (17K)  •  1954

Related videos

21:15
Analysis of main ideas and review of book, in animated format
9K    217
Oct, 2019
Spoiler: When analyzing statistics in advertising or the news, we need to check “sample size.” The sample size should be representative of the population, randomly selected, and large enough. A sensational headline can easily be drawn from a small and biased sample. We should always inquire about the size of the sample vs. the entire population when faced with a claim like “users of this toothpaste report 40% fewer cavities.” The reviewer notes that Huff’s book is free of complex mathematical equations and is accessible to all readers.
5:48
Top seven lessons from the book, explained over video footage
4K    66
Jun, 2022
Spoiler: One way to be more discerning when confronted with statistics is to “define the definition.” The definition of “family” or “school,” for example, may be different in different surveys. Charts and graphs are not to be trusted because they are easily manipulated; for example, the starting point of a graph can be changed to make the results look drastic. Beware of percentages; a claim of a “ten percent increase in sales” could simply mean an increase from ten sales on Monday to eleven sales on Tuesday.
1:34
Some core ideas of the book, explained in animated format
438    9
Jan, 2023
Spoiler: The use of averages, sampling errors, and misapplication of statistics are some techniques that are used to manipulate numbers. Misleading graphics, biased surveys, and faulty polls are all employed to sway public opinion. Huff provides tools for recognizing when statistics have been misapplied. He also outlines ethical principles that should be used when applying statistics and gives examples of manipulations from the past. In addition, the book discusses the power of statistics and the importance of understanding them.

Follow the author

Darrell Huff

Ask Albert:

Rate the book