I’ve seen this referenced a lot, especially with psychology, but can someone explain what exactly it means? How does correlation not imply causation? Sometimes, does correlation ever imply causation?
In: 160
Two things can be correlated but be completely unrelated and coincidental.
For example the [the number of pirates in the world is inversely correlated to global warming.](https://img.buzzfeed.com/buzzfeed-static/static/enhanced/webdr05/2013/4/9/15/enhanced-buzz-1129-1365534361-10.jpg?downsize=700%3A%2A&output-quality=auto&output-format=auto)
Obviously getting rid of pirates didn’t cause global warming 😉
Correlation never implies causation because that’s an *assumption*, and scientifically, we don’t want to be making assumptions. We want to dig into the stats and see if the causation is real.
Realistically, sometimes we can make an assumption. If your co-worker Mike only wears shorts on sunny days, assuming Mike is wearing shorts because it’s sunny is a pretty fair assumption to make. But you could also assume that Mike put on shorts this morning, forcing the weather to be sunny.
Correlation just means two things are related, such as ice cream sales and burglary, but one is not causing the other, they both just happen to occur more when it’s warmer out
If something is the cause of another thing, it’s no longer just correlation, and now you have a dependent and independent variable relationship
The way I explain it is… people with older children die sooner, but in MOST cases older children do not cause their parents to die. Just that older children and parental death occur together but are unrelated
Just because 2 things (A and B) happen at the same time (they are Correlated) it does not necessarily mean that A is causing (Causation) B.
E.g. Someone notices that when they see lots of umbrellas outside (A) it is also raining (B). But we won’t not say that umbrellas cause it to rain. The causation is actually there other way around in this example.
Probably better to say that correlation alone doesn’t PROVE causation. However, it certainly is a big clue that should be investigated.
Correlation means that two variables happen to have a statistical relationship. As one goes up, the other goes up. Or, if one goes up, the other goes down. But that relationship might be a complete coincidence or they might both be caused by the same underlying factor but are not directly related at all. For example, ice cream sales and shark attacks are very highly correlated in the US. But that doesn’t mean eating ice cream makes you more likely to be attacked by a shark.
Ice cream sales and shark attacks are highly correlated. Is one causing the other? Obviously not. It’s summertime and people are outside more and going to the beach more, hence more chance for both ice cream sales AND shark attacks. So there is a correlation, but ice cream sales and shark attacks are not causing each other. Banning ice cream will not have an effect on shark attack frequency. A lot of people will mistakenly see that two things seem to be related, and will try to “ban ice cream” as a solution to a problem when that’s not going to work because there are other major drivers at play.
This is different from a correlation that IS due to causation, such as: shark encounters are increasing, while water temperatures are also increasing. We know through research that climate change and rising ocean temperatures are affecting where sharks can live and this is making them more common near beaches where you wouldn’t normally see so many sharks.
Correlation does not **”imply”** causation is too strong of an assertion – it very often does certainly *imply* causation.
The phrase I’m more familiar with being used, especially in psychology, is correlation doesn’t equal causation or correlation doesn’t mean causation.
> but can someone explain what exactly it means?
It just means that sometimes two things can be correlated by random chance *or* that two things can be correlated due to a hidden variable, or that the cause and effect is backwards.
> Sometimes, does correlation ever imply causation?
Absolutely, there are literally millions of examples. The more massive an object is, the stronger its gravitational field. The gravitational field is correlated to the object mass.
In psychology, the personality of identical twins shares a higher degree of correlation than that of normal siblings. The implication being that genes cause personality to some extent.
You have two things A and B
Correlation is when A and B both happen at the same time. They both go up at the same time, and they both go down at the same time.
Causation is when there is a link between A and B, in that one causes the other.
But sometimes there is no link between A and B, it could be a complete coincidence that they rise and fall together. So there is no causation.
Latest Answers