I read this on the Writers’ Digest Newsletter today:
September: Worst Month of 2010 for Bookstores: According to stats from the U.S. Census Bureau, September sales were down 7.7 percent. This followed a 6.5% drop in August.
You’d conclude from this information that brick and mortar bookstores are doomed. That if things continue at this rate they’ll all be out of business in a year. The U.S. Census Bureau says so, right?
This isn’t quite the case, however. I remember reading (though I cannot find the source to cite) that August and September are traditionally slow months for book sales. It’s past the summer reading blitz and too early for Holiday buying to lift the numbers. A comparison between August and September 2009 to 2010 would give you a much more accurate picture of the health of the bookstore industry.
In a better example, [e-reads] uses statistics more carefully in this article about the growth of e-book sales, though I’d have liked them to include a comparison of the 39 million dollars of e-book sales to print sales for the same time period.
Poor statistical reporting leads to massive misunderstanding and misinterpretations. Take Amazon’s announcement that e-bo0k sales had exceeded hardcover sales. Talk show hosts quickly jumped on the drama train, reporting that traditional print books were on the ropes. But as the blog “Me and My Kindle” observed, that wasn’t what Amazon’s statistics implied even if you read them correctly.
We have so much information coming at us everyday it’s impossible to do the fact checking necessary to uncover all the misreporting, let alone the deliberate misleading use of statistics. Always look at the source of the info coming at you, and ask yourself what their bias is. Is their interpretation (and extrapolation) of the data valid? Ask what information is missing. (It’s harder to spot missing contradictory info than it is to spot erroneous data.)
Two books that can help you develop greater clarity of perception: How to Lie with Statistics by Daniel Huff and Damned Lies and Statistics by Joel Best.