Normal Distribution

mechanism

A normal distribution puts many observations near the mean, not necessarily at it. When a person must be average across several independent features simultaneously, that exact all-round averageness becomes increasingly unlikely.

A class can have an average mark of 42 even when not one student scored 42. The centre of a distribution can describe the group perfectly while describing no individual in it.

The peak is a neighbourhood, not a person

In a normal distribution, observations crowd around a central value and thin out toward the extremes. IQ illustrates the structure: the bell curve is centred at 100, with 15 points marking one standard deviation. But density near 100 does not imply that most people have exactly 100.

The distinction grows sharper across multiple dimensions. A face can be near-average in the width of the eyes, length of the nose, shape of the jaw, and many other features, yet the chance of landing exactly at the mean on every feature falls quickly. The statistical average is therefore often a useful coordinate without being a common specimen.

E1 E2

Where it shows up

The student nobody met

An average score summarizes the class, but treating it as the score of a typical student confuses a calculated centre with an actually observed case.

The hundred-point centre

IQ uses 100 as the midpoint of a normal curve and 15 points as one standard deviation. That locates scores relative to the population; it does not turn 100 into a complete portrait of an ordinary mind.

The elusive average face

People may cluster near the mean on each facial measurement while very few are jointly average on all of them. This is the statistical wrinkle beneath the Averageness Hypothesis of Beauty.

Not every crowd makes a bell

The lesson applies only after a variable has been shown to be approximately normally distributed; the IQ example supplies such a case, not a license to assume every human trait follows the same curve. And even for a bell-shaped variable, “near the mean” and “exactly average” answer different questions.

Replace the average person with intervals

The next time a report describes an “average person,” ask for the spread around the mean and list the separate dimensions being averaged. Treat each as a range, then check whether the supposed representative person would need to occupy all those ranges at once.