[wpdm_file id=12 description=“true”]How to Lie with Sta­tis­tics is a book writ­ten by Dar­rell Huff in 1954 pre­sent­ing an intro­duc­tion to sta­tis­tics for the gen­er­al read­er. Huff was a jour­nal­ist who wrote many “how to” arti­cles as a free­lancer, but was not a sta­tis­ti­cian.

The book is a brief, breezy, illus­trat­ed vol­ume out­lin­ing com­mon errors, both inten­tion­al and unin­ten­tion­al, asso­ci­at­ed with the inter­pre­ta­tion of sta­tis­tics, and how these errors can lead to inac­cu­rate con­clu­sions. In the 1960s and ‘70s it became a stan­dard text­book intro­duc­tion to the sub­ject of sta­tis­tics for many col­lege stu­dents. It has become one of the best‐selling sta­tis­tics books in his­to­ry, with over one and a half mil­lion copies sold in the English‐language edi­tion, even though the mon­e­tary exam­ples have become dat­ed because of infla­tion.[1] It has also been wide­ly trans­lat­ed.

(tekst: http://en.wikipedia.org/wiki/How_to_lie_with_statistics )

Themes of the book include “Cor­re­la­tion does not imply cau­sa­tion” and “Using ran­dom sam­pling”. It also shows how sta­tis­ti­cal graphs can be used to dis­tort real­i­ty, for exam­ple by trun­cat­ing the bot­tom of a line or bar chart, so that dif­fer­ences seem larg­er than they are, or by rep­re­sent­ing one‐dimensional quan­ti­ties on a pic­togram by two‐ or three‐dimensional objects to com­pare their sizes, so that the read­er for­gets that the images do not scale the same way the quan­ti­ties do.

The orig­i­nal edi­tion con­tained humor­ous illus­tra­tions by artist Irv­ing Geis. In a UK edi­tion these were replaced with car­toons by Mel Cal­man.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.