Have you heard the story about the guy who went straight to the top, without any setbacks along the way? He enjoyed one triumph after another, each conquered mountain taller than the previous peak.

Of course you didn’t hear that story. It’s obviously phony. Life is rarely so neatly navigated, so.tidy.

Neither are numbers, and to drive home the point I’ve taken the bold step to give it a name, Baron’s Theorem of Too Tidy Statistics: If a statistic sounds too tidy to be true, it’s likely hokum.

Case in point: I once heard a gentleman talking to a group about the impact of Baby Boomers on advertising patterns. Noting that the Boom spanned the years 1946 through 1964, the speaker said that the number of babies born increased each year for the entire period. In other words, there were more babies born in 1947 than in 1946, and more in 1948 than in 1947, and so on until 1964.

We all hear Boom statistics so frequently, but I had never heard that little piece of intrigue. The assertion seemed a bit “too tidy” for my tastes. It didn’t take long to see that it was quite a bit messier than portrayed. In 1948, just the third year of the Boom, the number of births dipped from the previous year.

Lest you suspect that was the only exception, six other times over the next 16 years, babies arrived in smaller quantities than the previous year. For the record, there were just fewer than 76 million people born during the 19-year period, or four million annually.

For the past 15 years, the average number of births in the United States has been about four million annually, ranging from 3.9 million to 4.1 million. Those figures approximate the Boom years, but keep in mind that the country’s population has more than doubled since the start of the Baby Boom, from 140 million at the start of 1946 to 291 million today.

A corollary to Baron’s Theorem of Too Tidy Statistics is Baron’s Rehashing of Other People’s Fine Analysis of Numerical Nincompoophood. I have also dubbed this “The Trouble With Double Tenet”: If it includes the phrase “doubles every (insert space of time here)”, then it’s usually hype or hyperbole.

To begin his book Damned Lies And Statistics, Joel Best highlights the peril of “double” references. He points to a dissertation written in 1995 by a student who stated, “Every year since 1950, the number of American children gunned down has doubled.”

Best retraced the stat’s steps to a 1994 publication of the Children’s Defense Fund.

“The CDF’s The State of America’s Children Yearbook–1994 does state: “The number of American children killed each year by guns has doubled since 1950.”

As Best went on, “Note the difference in the wording–the CDF claimed there were twice as many deaths in 1994 as in 1950; the article’s author reworded that claim and created a very different meaning.”

In addition to other potential problems, including whether the word “children” covers the same age range over that span, Best notes that the U.S. population grew by 73 percent from 1950 to 1994, so “we might expect all sorts of things” to increase accordingly.

The figure furnished by the student, said Best, “may be the worst-that is, the most inaccurate-social statistic ever.”

In fact, if taken literally, the number of homicides would have climbed to 8.3 billion in 1983-about double the world’s population at that time-and to more than 35 trillion at the paper’s writing in 1995.

So, be on the alert next time you encounter a claim that something is doubling every month, or every year-whether it’s the volume of dog clothing sales or the number of cellular phones in the United States.

Often, it’s simply a matter of someone making casual use of the word “every.” Instead of “every year,” maybe the overall number has doubled over the space of five or 10 years.

Go to www.google.com and type in “doubles every year” and see what happens-nearly all of the references are to technology and the Internet, and their alleged ability to double the quantity of information, data, and users annually.

Perhaps that is possible, at least in theory, in some areas of high-tech applications. But for the most part, it’s drivel. Do the math, and you’ll see just how untidy things can get.

One last note: Be leery of claims that something is the “fastest-growing” this or that. The smaller something is, the easier it is to grow. Glean the raw data behind the percentage change and see what story unfolds, and whether you even have a story.

Related Posts:
My Percentage / Percentage Point Primer & Plea For Apolitical Math
Red Flag: Palin’s Math Inflation Problem

Leave a Reply

Your email address will not be published. Required fields are marked *