The 'C' Words: Correlation Doesn't Equal Causation
From a young age, we spend a lot of time asking, "Why?"
We're always looking for causes and reasons, and it's something that doesn't really leave us as adults.
But sometimes we can look too hard for those reasons.
That's when we can fall into the trap you might know as the "false cause fallacy" — or by the words, "Correlation is not causation!"
It's a warning that should be inked into the minds of debaters and politicians, scientists and statisticians — but even they don't always avoid the trap.
These two "C" words have quite different meanings. Correlation is a measure of how closely two observations are related. Causation describes one thing that causes another.
They can be confused — sometimes purposely, often by accident. Just because two things tend to occur together doesn't mean that one causes the other.
Take this example: In the summer, people eat more ice cream, and they get more sunburns.
If we plotted these two things on a graph, we'd see a positive correlation: at the time of year when ice cream eating increases, so do cases of sunburn.
So there is a correlation between them. But we can't say that eating ice cream is the cause of sunburns. There is, of course, a third factor — hot weather — that drives both the increase in ice cream consumption and the increase in sunburns.
But the examples are not always this obvious. And while a strong correlation might mean causality, there are other possible explanations too — like a third factor such as in the previous example, or simply random chance.
This is why a good scientific study will usually note only a correlation between, say, certain foods or certain habits and developing a specific disease — because causation is very difficult to prove.
So how can we avoid the correlation-versus-causation trap?
Statistics expert Tyler Vigen says we should think critically about statistics we see, and examine the figures and methods for ourselves to see if they justify a causal relationship.
And perhaps we should keep asking, "Why?"