Statistical Tricks and Traps: An Illustrated Guide to the Misuses of Statistics

by Ennis C. Almer and Louie Joseph Cantal

Book Overview

First Published in 2017. Routledge is an imprint of Taylor & Francis, an Informa company. This description may be from another edition of this product.

• Format: Paperback
• Language: English
• ISBN: 188458523X
• ISBN13: 9781884585234
• Release Date: September 1999
• Publisher: Pyrczak Publishing
• Length: 58 Pages
• Weight: 0.50 lbs.
• Dimensions: 0.2" x 8.3" x 10.7"

Customer Reviews

What stats is the other guy using?

When I started to read this book I realized just how much we're saturated with stats every day.I loved the example that showed how reporters can use government statistics to arrive at outrageous headlines like: "The Shocking Increase in Cocaine Use."Or what it means when educators say that all children need to perform "at grade level." Do you want your child to do so?Do you run banner ads on your web site? If so, how do you count the number of your visitors to reach the cost you charge?Are you the biggest e-store in your specialty area? How did you come to that conclusion?Do you remember any of the statistics you learned in school? Do you ever have use for them?Well, this nifty little book gives great examples of how anyone can use statistics to prove their case.But, what I really liked the most is that it's a quick read. Even if you're not a statistical genius, you'll easily grasp the various examples. It's short, to the point, and it's relevant.Full of charts, graphs, illustrations, and cartoons the book covers everything you'll ever need to know about how statistics are used in the every day world. Some of what's covered includes: => using graphs to prove your point<p> => how percentages can lie<p> => when is average NOT average?<p> => definitions that make your case<p> => why making assumptions can make you fail<p> => associations & causation - how do they link?<p> => confounding -- two or more ways to draw the same conclusion<p> => sampling & how to make it work in your favor<p> => measurement - can you get the same results more than once?<p> => the big picture - how to use many smaller studies to reach your conclusion<p>Am I telling you to skew numbers to prove your point? No, but you should be aware of how others are doing just that. In the long run it might save your sanity as well as your pocketbook.<p>My only negative comment -- the price could be a little lower.

How to avoid twenty-seven major statistical misuses

In Statistical Tricks And Traps: An Illustrated Guide To The Misuse Of Statistics, Ennis Almer shows how to avoid twenty-seven major statistical misuses as illustrated when reporters misinterpreting government statistics to arrive at misleading headlines such as "Shocking Increase in Cocain Use"; or expecting all children to perform "at grade level"; or major corporations grossly misleading readers with full-page ads; or governmental organizations and polling organizations manipulating and misrepresenting numbers born out of surveys. Statistical Tricks And Traps is one of the most "reader friendly" presentations on statistical computations and presentations every publishers and is ideal for the non-specialist general reader seeking to understand and be prepared for the abuse and misuse of statistics in journalist, corporate, or governmental presentations.

Definitely worthwhile, but pricey

I first saw this book reviewed in "Skeptical Inquirer." As I'm a firm advocate of healthy skepticism, I frequently read books reviewed in that periodical. (And, God knows, you won't find those books also reviewed in more "mainstream" journals!)The authors start by recommending that students studying statistics read and absorb the information in the book. I'm not statistician but, were I to become one, I could see myself becoming a tad arrogant about my craft. (My apologies to true, ethical statistics professionals.) So I concur that those studying statistics, with professional aspirations to the field, should read it. Indeed, there's one chapter on the subject of a student's having designed a statistical study so as to reach the conclusion he had determined the professor wanted; he later acknowledged that and confessed to the professor. (One of the points of that chapter was that any study should be replicated--as the professor had done, thereby refuting the bogus findings--to confirm or deny the findings of the original study.) So, despite the alleged empiricism of the practice, statisticians too are human.There was an article in the Washington Post about a year ago on how statistics are abused--made up--to satisfy various ideological needs. As I work with many a "left wing" organization, I frequently see at best dubious stats used to make an ideological point--at the expense of truth. And right wing groups have been even more malicious toward truth, in some even more transparent ways. Challenges to these phony statistics are most often laughably easy. The credibility of the causes that the contrived stats were meant to foster, though, suffer because of the challenges. Then there's the commercial stats, used to sell any number of products. The book covers these abuses, in a graphic but not frivolous way. (I don't know about you, but I get tired of the "idiot's guide" type of books that try to be humoruous. This book is not of that genre.) There are chapters that are humorous only in that they are likely to have the reader blushing at having been gullible to "studies," samples that are too small to be "statistically significant;" questions asked to those who didn't even understand the questions; idiotic conclusions demonstrating more the bias of the "investigator" than the validity of his or her findings, and many more. It's happened to us all.Then there was an article in a recent issue of the Post which headlined an elaborate equation with which alleged statistical pros claim to predict election results. There is a chapter in the book dedicated to that subject too. (Incidentally, a few days after the article was published, there were letters to the editor questioning that equation and the "pros" who use it, saying, among other things that, if it's true, then the election result has nothing to do with the campaign, and that "any Joe" could have predicted what this elaborate equation supposedly had predicted