One of the most common mistakes made when science is not in your comfort zone, is to confuse correlation and causation.
Correlation describes an association between types of variables: when one variable changes, so does the other. These variables change together: they “covary”. A strong correlation or “covariation” might indicate causality – however, it is important to remember that covariation isn’t necessarily due to a direct or indirect causal link. There could be another explanation: it may be the result of random chance (where the variables appear to be related, but there is no true underlying relationship), or there may be a third variable that makes the relationship appear stronger (or weaker) than it actually is.
Causation means that changes in one variable directly bring about changes in the other; meaning that there is a cause-and-effect relationship between two variables. The two variables are correlated with each other and there is also a causal link between them.
Let’s look at an example, of two variables that “covary”:
“Ice cream consumption rises in the summer”
“Heart attacks rise in the summer”
Just because these two variables both increase in the summer, does that mean that one is directly caused by the other? To determine the answer to this question (i.e. to determine ‘causation’), many more questions must first be asked, and many more considerations must first be taken into account. For example:
- Does ice cream cause heart attacks?
- Does heat cause increased ice cream consumption and heart attacks?
- Do ice cream promotions in the summer cause increased ice cream consumption?
- Does increased physical activity in the summer cause heart attacks?
- Do external factors we don’t know about cause these increases?
While it can be tempting to assume a “causal link” between two variables that “covary”, it is important to never jump to conclusions or state assumptions as fact – especially when reporting on science and health issues. Instead, train yourself to keep an eye out for these types of “patterns” or “covariations” in your investigations – and then challenge yourself to ask the type of questions that determine whether such “covariations” are in fact just correlated, or whether one may actually be caused by another.
In fact, setting yourself such a challenge can be similar to setting a Hypothesis – to prove or disprove an assumption you have – and can make for the basis of an extremely interesting and unique story!