Correlation and Causation Explained

Paul Schingle's image for:
"Correlation and Causation Explained"
Image by: 

In psychology, or in any of the sciences, one must distinguish the difference between causation and correlation.  Some correlations are causative.  When the sun comes out, it gets lighter, or it gets warmer.  This is a causative correlation.  The first act, the sun rising, causes the second act, it gets lighter and warmer.  But, sometimes, two events may occur correlatively without one event causing the other.  A common example taught in most introductory psychology classes illustrates this point.

While this sounds silly and extreme, this is also a statistically accurate statement.  There is a high correlation between the occurence of rape and the sale of ice cream.  Now, if this were a cause-effect correlation, any of the following could be true:

One) A perpetrator of rape may have an uncontrollable desire to have ice cream after committing the dastardly deed.  Two) A potential rapist eats ice cream and then has an overwhelming desire to commit the felonious act.  Three) A potential victim of rape eats ice cream, then places herself in harm's way or Four) After a vicitm has been violated, the only thing that can calm her down is to have some ice cream. 

Now, of course, every one of these scenarios is ridiculous.  And, why are they ridiculous?  Because, there is no cause-effect reason to explain the correlation between the occurence of rape, and the sales of ice cream. Yet, statistically, the correlation does, in fact, exist.  There is a high positive correlation between rape and ice cream sales.  The correlation is just not causative.

The classic explanation in the above mentioned (hypothetical) beginning psychology class is that Summer is the link of the two (non-causative) events.  Rape occurs more in the Summer (possibly because potential rapists are out of the house more, or more potential vicitms are outside or for whatever reason).  This is just statistically true.  Also, it is statistically true that ice cream sales increase in the Summer, probably for obvious reasons.  The bigger point of the issue is, though there is a correlation between rape and ice cream sales, the correlation is not a cause-effect relationship.

In a nutshell, that is the difference between causation and correlation.  While similar, they aren't the same thing.  Two things can correlate, statistically, but that doesn't mean there has to be a causation for the correlation. 

More about this author: Paul Schingle

From Around the Web