OK, the first thing is incredibly simple: How to determine where correlation = causation. You must firstly be able to cross-reference the citation and give conclusive proof that there is only 2 variables that are..well...varying. You need 1 independent variable that you control to affect the dependent variable.
So, if there is a study where there are multiple variables, and it claims to be one of them specifically, it can be disregarded. However, if there are multiple studies, where only 1 variable is constantly changed, and the rest are controlled, then it can make an accurate conclusion. An example:
If 1 study on whether allowing weapons to be legal correlates to a large murder rate, showing difference between London and LA, and saying that the legal guns is the cause of the murder rate, would be invalid.
If there was a comparison the same as this, but done with a hundred cities between America and Britain, then it would be a stronger conclusion, but still invalid to make the claim it is solely legal guns (claims could be made about culture).
If there were many countries doing comparisons of towns against other countries, including many (at least 3) countries from both sides, then the conclusion will be valid (unless there is a more accurate conclusion. For example, comparing Paris, Venice and London with Tripoli, Zawiya, and somewhere else, would be invalid as there is a civil war going on).
I hope this clears up when correlation can equal causation. Remember it does not always mean it does, but it does many times. As usual, peer reviewed journals or studies are the best source for the information.
No comments:
Post a Comment