How to Lie with Statistics

Journalist Darrell Huff, a freelance writer of how-to articles rather than a statistician, explains for general readers how statistics are misused and misread. He covers sample studies, tabulation methods, interview techniques and the way results are drawn from figures, along with themes such as correlation not implying causation and random sampling. Huff shows how charts can distort, for example by cutting off the bottom of a bar or line chart to exaggerate differences, or by using pictograms whose images do not scale the way the quantities do. The original illustrations were by Irving Geis. In the 1960s and 1970s many college students met statistics through it, and the English edition has sold over one and a half million copies.

Bill Gates said it feels more relevant than ever despite a few dated examples, and Morgan Housel said it makes readers laugh and grow skeptical of nearly everything in a single sitting.