Seeing the world in numbers: the eye of statistics
A hundred scores, a thousand heights. Looking at each one makes your head spin. But what if just two numbers could sketch the whole picture? Where the middle sits, and how spread out things are. With these two, the data shows itself at a glance. This is the eye of statistics, and the first eye an AI uses to look at data.
Where is the middle? The average
Data is scattered as points.
With many points, none of it fits in one look.
So the very first question is this.
Where do these points roughly sit?
Knowing that middle
catches the whole shape in one number.
The most common way to find the middle
is the average: add all and divide by the count.
It is the spot where the added weight balances.
Tap points to add them one by one. Each new point shifts the middle average marker to the balance point.
The more points you add, the more the marker
followed the balance point of the whole.
If points pile up to one side,
the average is pulled that way too.
So the average tells well
where the data roughly is.
Yet the average alone
misses one thing.
The same middle can come in different shapes.
How spread out? The scatter
Here are two groups.
Both have the very same middle value.
Yet one group huddles tight at the middle
while the other spreads out wide.
Does the same average mean the same data?
Not at all.
The tight side is all rather alike,
while the wide side is uneven.
This degree of scatter is called spread.
Tap the two groups by turns to compare. The middle value is the same, but the spread bar lengths differ.
The middle value was the very same,
but the spread differed at a glance.
The tight group had a short bar,
the wide group a long bar.
So average and spread are a pair.
Seen together, the data shows truly.
Where the middle is (the average),
and how far it scatters around it (the spread).
These two numbers are the basic eye of statistics.
One faraway point, the outlier
The points sit clustered cozily.
The average is right in their middle.
But if one point comes in from far away,
what happens?
Such a point set apart from the rest
is called an outlier, a stray value.
Since the average adds all and divides,
even one faraway point pulls it along.
A single point can shove the middle over.
Tap the faraway point to add it to the group. One point yanks the average marker sharply that way.
Just one point came in,
yet the average swung that way.
Even though the rest stayed put.
So looking at the average alone can fool you.
Most are gathered in one place,
but one stray value rocked the middle.
When you look at the spread too,
it signals the points scatter wide.
That is why you watch the two numbers together.
Lots of data into two numbers
Now there are very many points.
Looking at each one is too much.
But we have learned two numbers.
The middle value and the spread.
With these two, be it a hundred points or a thousand,
you can summarize the whole in brief.
The middle is about here, the scatter is this much.
This knack of shrinking many into few
is the starting point of how AI handles data.
Tap the compress button. Many points shrink into just two numbers, the middle value and the spread.
A great many points
shrank into just two numbers.
Of course, shrinking loses something too.
The fine detail of each point fades away.
Still, with only two numbers
you can grasp the big character of the data.
Where it gathers and how far it scatters.
Keeping just the essence and summarizing like this
is the most basic eye for looking at data.
Let's wrap up
Gathered on one line, it is this.
Lots of data gets summarized into few numbers.
The average is the middle value, added and divided,
and the spread is how far the points scatter.
Even with the same average, a different spread
means utterly different data.
A faraway outlier rocks the average,
so always look at the two numbers together.
The middle and the scatter, that is the eye of statistics.
Tap the key points in order to review. (the average is the middle -> spread is the degree of scatter -> same average with different spread means different data -> an outlier rocks the average)
Now you have the first eye
for seeing data in numbers.
The dizzying crowd of points
grew clear through two numbers, the middle and the scatter.
This power to summarize many and grasp the essence
is where an AI begins to see the world.
Next, how do we read the relationships
between these numbers?
Let's keep exploring together.