Cover image for Statistics you can't trust : a friendly guide to clear thinking about statistics in everyday life
Statistics you can't trust : a friendly guide to clear thinking about statistics in everyday life
Campbell, Stephen K. (Stephen Kent)
Publication Information:
Parker, CO : Think Twice, [1999]

Physical Description:
270 pages : illustrations ; 23 cm
General Note:
Includes glossary and index.

Some parts of this book appeared in Stephen K. Campbell, "Flaws and Fallacies in Statistical Thinking" (New Jersey: Prentice-Hall, 1974).
The last shall be first -- Measurements: valid, veiled, and overblown -- Cheating charts -- Accommodating averages -- Air, gravity, and variation -- Puffing up a point with percents -- Careless comparisons -- Jumping to conclusions -- Slapdash probabilities -- The mysterious sampling factory -- Relationships: causal and casual -- Beyond plain as day -- Significant others -- Forty awful examples.
Subject Term:
Added Author:
Format :


Call Number
Material Type
Home Location
Item Holds
HA29 .C255 1999 Adult Non-Fiction Non-Fiction Area

On Order

Reviews 1

Booklist Review

Often as we hear about "lies, damn lies, and statistics," most of us occasionally get fooled by the numbers in news stories, sales pitches, even family arguments. Campbell, a University of Denver professor emeritus who taught statistics for more than 30 years, aims to make us less gullible with suggestions on distinguishing valid from invalid statistics and sound from confused or manipulative use of valid statistics. His approach is humorous, taking on "cheating charts," "accommodating averages," "careless comparisons," and other flaws of logic or mathematics or both in brief, punchy chapters that include samples of statistical confusion as well as comic illustrations by cartoonist Mark V. Hall. Other topics covered include measurement, variation, percents and probabilities, sampling, causality, contingency factors, and regression analysis. Campbell closes with "Forty Awful Examples," whose errors the reader can identify with help from a glossary that lists common statistical errors as a result of "simple carelessness or inept presentation," "confusing something with something else," "opportunistic selection," or "withholding/overlooking of important information." A useful, entertaining overview. --Mary Carroll