Correlation vs. Causation

By Deane Barker

Just because two numbers move together (are correlated) that doesn’t mean one number caused the other. Even if causation exists, there’s no telling what caused what.

Examples

For example, if the datapoints of (1) the amount of household income, and (2) number of books in the household move together, this does not prove they’re related, or that one causes the other. Some may say, “This proves that people who read more become wealthier,” thus implying that more books causes wealth. In reality, it may just mean that wealthier people have more income and space in their house for books, or that they went to college, or any number of other scenarios.

I saw a meme on Reddit that said this:

Headline: “Women Who Own Horses Live Longer”

Implied correlation: horses make you live longer

Reality: if you own a horse, you can probably afford health insurance

This is item #21 in a sequence of 87 items.

You can use your left/right arrow keys to navigate