The term “on average” denotes the usual amount of something. However, using “average” in presentations, media and more can be terribly misleading and it is up to the data driven executive to make proper (and ethical) use of this calculation.
George Orwell’s poignant tome 1984 presents a dystopian future where a surveillance police state has control over the populace in just about every regard from health, food, speech, and education. Large telescreens mounted on street corners and in the home, blast statistics at denizens regarding how good life is under Big Brother; “Day and night telescreens bruised your ears with statistics proving that people today (on average) had more food, more clothes, better houses, better recreations—that they lived longer, worked shorter hours, were bigger, healthier, stronger, happier, more intelligent, better educated, than the people of fifty years ago.”
Statistics can definitely mislead and there are few phrases that bruise the ears more than “on average”. In fact, our minds should be automatically alerted to pay special attention whenever this terminology is used in speech or prose.
An egregious example of the use of the term “average” comes courtesy of a freeway billboard near a local hospital in San Diego. In a display of courage, the billboard proudly pronounces that their “average ER wait time is 30 minutes”.
Now, on the whole, this wait time may sound like a positive attribute, especially considering national papers often highlight emergency horror stories of up three hours (or more) to receive care. And ER wait times have increased over the past decade, so perhaps thirty minutes is really something to trumpet.
But does this statistic really tell the whole story regarding speedy hospital service? Some questions to ponder:
- Considering 24 hours in a day – is the wait time from 12am-6am ten minutes, while during the day it’s an hour? We just don’t know.
- Does “30 minutes wait time” mean wait time to see a doctor, diagnostics or simply just to see the triage nurse?
- In the statement “average ER wait time is 30 minutes” – an automatic response of the reader/listener should be “compared to what”? In fact, the other local hospital a few miles away may have an “average” wait time of just 20 minutes…
So if statistics sometimes obfuscate, the solution according to author Darrell Huff is to fight back against such abuse. He says that when it comes to statistics, executives should always be on the hunt for bias in sources and samples, and ask “what’s missing”. Sometimes what is presented isn’t as important as what’s hidden.
Lastly, Huff says we should judge statistics with good old fashioned common sense. As in suppose you happen to be in a room of ten people. When Warren Buffett enters the room, are you really “on average” a billionaire? The answer is yes, however it’s too bad your bank account will not concur.
