Darrell Huff's How to Lie with Statistics - Book Report/Review Example

Darrell Huff's How to Lie with Statistics Book Summary As reviewed by The Atlantic, How to Lie With Statistics is, “a pleasantly subversive little book, guaranteed to undermine your faith in the almighty statistic.” Truly, the book even though a bit dated, having been written in 1954 is timeless. A classic that persists to chase away false beliefs and enlighten those who are statistically naive. Whether you encounter statistics at school, at work, or in advertising, you'll definitely commit to memory its simple lessons. The witty and direct style exposes how media, advertisers, and government misinform their audiences through the misuse and abuse of statistics. Huff then explicates how the reader can see through the smoke and to get to what really lies behind the mirror. "There is terror in numbers," writes Darrell Huff. His book aims to decipher the terror that lies beneath the world of averages, trends, graphs, and correlations. Huff sought to break through "the daze that follows the collision of statistics with the human mind.” The book remains relevant as an awakening for people unacquainted to delve deeper into the nonstop flow of numbers pouring from Madison Avenue, Wall Street, and everywhere else; where someone has a point to prove, a product to sell or an axe to grind. Darrell Huff investigates the breadth of every popularly used type of statistic, explores such things as the tabulation method, the interview technique, the sample study, or the way the outcomes are derived from the figures, and points up the infinite number of dodges which are used to deceive rather than inform. "The secret language of statistics, so appealing in a fact-minded culture, is employed to sensationalize, inflate, confuse, and oversimplify," warns Huff. On the other hand, he said that we should not be terrorized by numbers. "The fact is that, despite its mathematical base, statistics is as much an art as it is a science." Synonymous to a lecturing father, he expects you to learn and ponder on something valuable from the book, and start applying it every day. Never be a sucker again, he cries! Deceptive Presentation of Graphs (Truncated or The Gee-Whiz Graph) Seeing graphs illustrating numbers if properly done are very helpful in interpreting and analyzing data. And yet, truly deceiving if completed in a fishy fashion If you want to show statistical data, clearly and quickly. Draw a picture of it. When a graph is constructed with a y-axis that is numbered from 1 to 100 without skipping a unit, Huff explained, "Your ten percent looks like ten percent—an upward trend that is substantial but perhaps not overwhelming. That is very well if all you want to do is convey information. But suppose you wish to win an argument, shock a reader, move him into action, and sell him something." Huff suggested you could magnify the statistical trend. "Chop off the bottom. Of course, the eye doesn't 'understand' what isn't there, and a small rise has become, visually, a big one." Clearly, as the graph truncates the lower portion, it exaggerates the magnitude of the movement. Huff continued, "Now that you have practiced to deceive, why stop with truncating [cutting off the bottom]? You have a further trick available that's worth a dozen of that. Simply change the proportion between the ordinate and the abscissa. It is a subtler equivalent of editing 'National income rose ten percent' into 'climbed a whopping ten percent.' It is vastly more effective, however, because it contains no adjectives or adverbs to spoil the illusion of objectivity. There's nothing anyone can pin on you." It’s modifying the scales so that the rectangles aren't square. To make a trend visually more dramatic to readers, use an uneven spacing on the along axes. In this scenario, "dramatic gains" appear less dramatic. In this scrupulous case, that could have been simply noted with a squiggle to point out the missing part - a distinctive technique used by publishers and writers to save space. In the circumstance pictured, most readers have a representation that the trend is positive, thus a favorable result. But if you strict to the truth, results are more truthful and thespian with evenly spaced scales on the axes and squares instead of rectangles as the units. A quote from Mark Twain perfectly characterizes the gee-whiz graphs, "There is something fascinating about science. One gets such wholesale returns of conjecture out of such a trifling investment of fact." The One-Dimensional Picture There's another neat trick - flaunting a one-dimensional value in two dimensions. Huff illustrates the use moneybags, one twice as tall as the other. Note in this picture, different sized money bags are used to demonstrate that American workforce earn twice as much as their Russian counterparts. The picture is seemingly representing one person having twice as much money as another. The trickery is visual; although the bigger bag is correctly twice as tall as the smaller one, the area you see in the picture is four (4) times larger. And the volume, whew, eight (8) times larger! This is done by doubling the height of the money bag as well as the width. Presenting a four-fold increase in area and thus a four-fold increase in the American wage. And of course bags are three dimensional, therefore, the American's bag can hold eight times as many coins as the poor Russian's. All from a simple doubling. As good as playing tricks, your eye will single out the volume or area and deliver an ambiguous notion. Quoted by Aaron Levenstein, “Statistics are like bikinis.  What they reveal is suggestive, but what they conceal is vital.”  Distributions (Mean, Median, Mode) Speaking of a group of values, "average" value as the sum of all the values come into picture. You might presuppose one thing but, on instance, someone with an hidden agenda will use something different. The usual assumption is the arithmetic “mean” - which is the sum of the numbers divided by the count. And there is another "average" that is occasionally used, namely the median. Once you take all the values and arrange them in ascending order, take the middle one, and you have the median. Hence, it means that half of the values were higher and half were lower. As an illustration of the above scenario, Duff, stated, “I live near a country neighborhood for which I can report an average income of $15,000. I could also report it as $3,500. If I should want to sell real estate hereabouts to people having a high snobbery content, the first figure would be handy. The second figure, however, is the one to use in argument against raising taxes, or the local bus fare. Both are legitimate averages, legally arrived at. Yet is obvious that at least one of them must be as misleading as an out-and-out lie. The $15,000 – figure is a mean, the arithmetic average of the incomes of all the families in the community. The smaller figure is a median; it night be called the income of the average family in the group. It indicates that half the families have less than $35,000 a year and half have more. Here is where some confusion about averages comes from. Many human characteristics have the grace to fall into what is called the “normal” distribution. If you draw a picture of it, you get a curve that is shaped like a bell. Mean and median fall about the same point, so it doesn’t make very much difference which you use.” A small amount of large values will distort the distribution and haul the arithmetic average up but will not have much effect on the median. Another parameter of a distribution that is worth mentioning is the mode. The mode is the value that occurs most often or the peak of distribution. If numbers are distributed truly randomly otherwise known as the classic bell curve, the average, median and mode will coincide. Otherwise, the three - mean, median and mode will separate, if the distribution is not bell-shaped. For Rex Stout, “There are two kinds of statistics, the kind you look up and the kind you make up.”  Points of View Even though many of the examples used in the book are enchantingly dated, the concerns and information are enduring. Statistics are prevalent with opportunities for abuse, from "gee-whiz graphs that add nonexistent drama to trends, to results detached from their method and meaning, to statistics' ultimate bugaboo--faulty cause-and-effect reasoning. Huff's tone is tolerant and amused, but no-nonsense. Even if you can't find a source of demonstrable bias, allow yourself some degree of skepticism about the results as long as there is a possibility of bias somewhere. There always is.” (Littleton, 2000). In book, “How to Lie with Statistics,” By Darrell Huff was very informational and interesting to read. I was carried and blown away on how people can really cheat on statistics. I do not mean to be prejudice, though not all statistics are misuse -- it is deplorable how can people have the guts to show something so good to be true on graphs and get away with it. ”Say you were standing with one foot in the oven and one foot in an ice bucket.  According to the percentage people, you should be perfectly comfortable.”(Bragan, 1963). This is what puffs my mind. For example, when Darrell Huff talked about “The Sample with the Built-In Bias,” he used a paradigm by using the class of 1924 of the Yale men. He proved how not all Yale men were included in the statistics of “if you go to Yale you will make a $25,111 figure.” Since not all of the Class of 1924 addresses were known, to check if they were also having the same figure. Like Huff said, “If all the Class of ’24 is included, that’s all right. If every tenth name on a complete list is used, that is all right too, and so is drawing an adequate number of names out of a hat.” Theoretically, Duff’s main point is, as long as you have some percentage of what you are trying to establish, you can confirm it no matter what little you have; and that it is fine. Thus, Huff said, “You can prove about anything you want to by letting your sample bias itself.” "There are three kinds of lies:  lies, damned lies, and statistics."  Mark Twain, References : 2000. Amazon.com, Inc. 15 Mar. 2007 . "Book Group Online - How to Lie With Statistics." 2000. Jelsoft Enterprises Ltd. 15 Mar. 2007 . "Lying With Statistics." 15 Mar. 2007 . Electronic Book   Basic Book   Reference Work   Republication Electronic Anthology   Author & Editor   Author & Translator   Organization & Editor Electronic Periodical   Journal   Magazine   Newspaper Huff , Darrell, How to Lie with Statistics . New York: Norton, 1993. "NCTM: News & Media: Does the Graph Tell the Story (Newsbulletin November 2002)." 2000. National Council of Teachers of Mathematics. 15 Mar. 2007 . Read More