Deming

You really need to be above average!

Photo by Stephen Phillips - Hostreviews.co.uk / Unsplash

You really should be above average? At the very least!

Isn't that what society expects of us? What schools teach?

Who wants to be average? Even worse, less than average?! Perhaps there are some who don’t care, but how many parents want to hear that their child is anything less than ‘above average’?

So everybody strives to be above average. And that’s when the irony really kicks in from a statistical perspective.

Do we really know what it means to be average?

We hear the term so often. The average person. The average lifespan. The average salary.

But, how often does anything actually fit with being that average?

Less often than we believe.

We rely on a concept that fails to account for the very individualism that makes us human, that allows for the natural variation that exists in every and any system or process.

Eye opening on this, and a fantastic read is Todd Rose’s The End of Average. Examples abound on the flaws of relying on the average, how treating things (people) as if they are the same, the average of some measure, causes much harm.

And no surprises. As those who know me, perhaps read my blog articles and posts (I hesitate to refer to them as social media posts), all sits nicely within Dr Deming’s System of Profound Knowledge (SOPK).

Jokes abound about the average

Dr Deming had a few jokes he would make when talk of average was raised, as do many statisticians.

“A man with his head in an oven and his feet in a bucket of ice, on average, feels fine.”

“The river is only four feet deep on average.”

Somebody like me prefers to be able to keep my head above water with my feet touching the bottom, so my height means I’m safe across this river! At least assuming the current doesn't sweep me away.

And not even a joke when we teach our children in school to ignore such limitations. Dr Deming referred to Dr Lloyd Nelson’s anecdote about his daughter who was graded poorly by a teacher:

The teacher asked the class: ‘What is the average family size in America?’ Lloyd’s daughter answered: ‘There’s no such thing as an average family.’

Can an average family size of 4.4 actually exist? Physically impossible!

But a mathematical answer that does not exist in reality and the actual correct answer given by a child. No wonder we destroy the natural curiosity through the prevalent school systems (I was going to suggest ‘traditional school systems’ but then worried ‘traditional’ was hard to define and too similar to ‘average’).

On a serious note, what does SOPK have us think about average?

If we limit ourselves to think of the average mathematically, we fail to understand and avoid the psychological harm it can cause.

Given at any moment in time half will be above average whilst the other half will be below. That’s statistical and randomly distributed. Suggesting a child is below the average might make sense mathematically, but damages them psychologically. For no reason!

The average does not take account of the variation inherent in the system, whether it is a class or manufacturing process. We need to appreciate that we are considering a system, that has interactions within and outside, and is not just about the individual parts. We need to understand variation to understand the system. There are people in this system, each are different, none the same, none exactly the average value, with their own minds, their own motivations. And part of all this, as interdependent as each other part of SOPK, we have knowledge, theory, curiosity to test scientifically.

In reality, the mathematical concept of average is only that, a concept. That same system will have a different average at another point in time (if calculated separately to the first group) because of the natural variation in their system.

I'll avoid getting into ranking students based on a test score. Only to say, what is the actual difference between somebody who scores 96% and 95%? Who has decided there is a difference, and by what method? Look like it would be part of the normal variation in a system, not a true difference.

Ban the average benchmark

Comparing anything, and especially when talking about human beings, to the average should be banned. We'll be below the average 50% of the time, and likely through no 'fault' of our own!

Accept there is natural variation within a system or process, use the correct scientific methods to identify this. Use these to work in improving the system and determine those that fall outside the system (either better or worse) and dedicate the specific approaches to help them (either keep performing better or bring performance to the system’s natural variation, which is a range).

Best, compare individuals only to themselves and not others, for their own improvement - accepting that actual improvement of a stable system will only come from a re-design.