Saturday, June 02, 2012

On Beyond Average

Even if you barely stayed awake in math or psychology or almost any class for that matter, odds are you have heard of the bell curve. It's a way of describing the phenomenon that, when sampling virtually any group, be it human or animal or mineral, the distribution is such that most will fall somewhere in the middle, tapering to fewer at either end. Doesn't matter what you are measuring: size of rabbits, weight of kid's backpacks, shoe size. In almost every case, the vast majority can be defined as "average" with a decreasing amount above or below that marker.

You see this reflected in almost everything around you. If you go to the store, you will find more shirts in medium, fewer in large and small, still fewer in XS and XL. Same goes for automobiles: most are mid-sized, with Smart Cars at one end and Chevy Tahoes at the other. And if the balance starts to shift, the theory still holds: the center moves with it, and the relationship stays intact. Yes, we are all getting fatter. But as the average weight increases, so too does the definition of what average is, and so the distribution plays out as per what you'd expect.

We also think of this as the rule of human achievement, with most of us solidly in the middle, and a decreasing number of superstars and poor performers as you look further out in either direction on the scale. Indeed, schools are designed with this basic assumption in mind, with classes and lesson plans catering to those in the middle, and a few for others who need remedial work or who can fly beyond the material. It's the same be it chemistry or music or gym.

But what if that wasn't the case? What if that so called "normal distribution" was not a report of what has happened, but rather a system that actually forces people to settle into it? That's the conclusion of researchers Ernest O'Boyle Jr., of Longwood University's College of Business and Economics located in Farmville, VA, and Herman Aguinis at Indiana University's Kelley School of Business, in their study with a slightly skewed pizza-box title, "The Best and the Rest."

O'Boyle and Aguinis surveyed four broad areas of performance: academics writing papers, athletes at the professional and collegiate levels, politicians and entertainers. They looked at 633,263 people involved in those endeavors, spread over five separate studies involving 198 samples, each surveying achievement in a group. And what they found, contrary to popular belief, was that a small percentage of superstars accounts for most of the performance, while the rest of the population is actually not average, but below average. When plotted they found not a normal "Gaussian" distribution, but a power law "Paretian" distribution. Put more simply, rather than the aforementioned bell curve, you find a sloping right triangle, with a few heavyweights on the far right, and the rest of us clustered to the left on the low end of the scale (Speak for yourself, you're saying, but the truth can hurt).

Take academic paper writing as an example. As the researchers explain, "A normal distribution and a sample size of 25,006 would lead to approximately 35 scholars with more than 9.5 publications. In contrast, our data includes 460 scholars with 10 or more publications. In other words, the normal distribution underestimates the number of extreme events and does not describe the actual distribution well." Or consider achievement in more popular fields, as defined by "Award nominations (Oscars, Emmys), expert rankings (Rolling Stone), and appearances on a best seller list." In every case, a few individuals stand out time and again (I'm looking at you, Meryl Streep) while most others rank their achievement in "People" mentions as opposed to actual accomplishments (can you say "Kardashian?).

The "why" of this anomaly is a question. Perhaps, the paper muses, it's because human achievement is often constrained by outside events, be it an assembly line, classroom or an election. Superstar craftsman, students or politicians can't shine, and so most get lumped into the middle, producing the bell curve we find so comforting. Remove the shackles, goes the reasoning, and more widget makers of whom we can be awed will rise to the top.

Food for thought for the future. In the meantime, you can certainly hide in the center of the curve, comfortable with the rest of us. Or you can throw off the fat part of the curve that is holding you back, move towards the right side, and call Steve Jobs, Bill Clinton, and Beyoncé your new homies. Just don't be bound by the numbers. Turns out Mark Twain was right: there are lies, dammed lies, and then there are statistics.


Marc Wollin of Bedford tries to be extraordinary in an average sort of way. His column appears regularly in The Record-Review, The Scarsdale Inquirer and online at 

No comments: