How to Lie with Statistics

Reference Resources – Knowledge is Power!

~submitted by Doug Murdock, Globe University-Sioux Falls Campus Librarian

I was wondering— How many of you have ever been in a conversation that began with “I bet you didn’t know that….” And then quoted some barely plausible statistic like this: I bet you didn’t know that between 1890 and 1953 that Oklahoma City had an average temperature of 60.2 degrees? Now on the surface of it, what does this number actually tell you? That Oklahoma City has a nice, very liveable temperature and would be a pleasant place to live on average? Not really; here’s what that number DOESN’T tell us. To come up with that average, on any given day the temperature could vary by as much as 130 degrees!

Another thing; in this season of politics, everyone on both sides has heard what I call a “tricky” kind of statistic; one that makes sense on the surface, but when you analyze it is full on nothing but bushwa. Here’s an example from the book. Just before the end of the Korean War, the Chicago Journal of Commerce ran a story that discussed the results of a survey of businesses about hoarding and price gouging that they conducted. Their comments, in essence, stated that two-thirds of the businesses were not hoarding or price gouging.

Now let’s look at this and “do the math.” They surveyed 1,200 companies and only 14% replied; that’s 168 out of 1,200. Of these, 9% (16) said they hadn’t hoarded or raised prices, and 5% (9) said they had! Their report was implying that 2/3 of the 1,200 companies surveyed (792) weren’t hoarding.See how statistics can mislead?! 

This book is short; only 144 pages, and with examples like these is a joy to read. Also, and most importantly, it is full of methods you can use to deconstruct misleading statistics, and to make sure yours are good. Come up to the library and have a look at this gem. You won’t be sorry!