Can someone Eli5 please.
In: 8
Correlation is a trend where something occurs and you see that hey. This other things tends to occur more or less. But it doesn’t mean that one actually causes this to occur.
My fav example. Correlation between margarine consumption and divorce. The more margarine consumed. The greater the divorce rate. https://www.tylervigen.com/spurious-correlations
Causation on the other hand. If I take 100 people. I believe that putting your hand on a hot stove will cause a burn. I put it to the test. Yes. 100 people burned their hands. And it was caused by the heat from the stove.
That being said, correlations are useful and necessary. We’ll defined variables/covariates and modeling. We can’t always conduct the cause and effect studies. But we can reasonably infer that there is a strong relationship between two (or more) variables.
Latest Answers