Question

One hears versions of the expression "correlation does not mean causation" so often - especially as a graduate student - that it can be easy to accept the statement as true without giving the matter some careful consideration.

Is it ALWAYS true that correlation does not imply (or suggest) causation? Does correlation NEVER imply (or suggest) causation? Are there any conditions under which we might have good reason to think that a causal relationship does exist on the basis of a correlative relationship? If so, what are those conditions?

Comment, please.

P.S. Usually, students who create problems for themselves with this prompt have a poorly-formed idea of causation. If you are not confident that you know what causation means, you might want to spend a few minutes looking it up before you jump into crafting your response.

Answer 1

Answer: Yes I feel that correlation does not imply causation. It only indicates a higher probability that two events have some type of relationship with one another. If one event is followed by another event it does not necessarily means that the second event was caused by the first one. There can be a relationship between two variables but causation is a different thing. For two variables to be interrelated it is not essential that one may be a cause for another. Causation means that one variable is a reason for the other to happen. There can be a correlation between things because of a third variable which may act as a real cause for causation of an event. For example the wind mills rotate faster, when air is blowing fast. Therefore there exists a correlation between fast wind and the fast rotation of windmills but this does not implies that fast rotation of windmill causes fast wind to blow. But it does not mean that correlation never implies causation as in this case it fast wind can be said to cause fast rotation of windmill. Therefore At times in one situation itself it is not easy to ascertain that whether correlation does not imply causation.