The Ice Cream Murders: Correlation vs. Causation

As students head back to school and we return to academia, it is a good time for all of us to refresh our scientific know how. You don’t have to have a Ph.D to understand important scientific fundamentals.

No really, it’s NOT that complicated. I swear I can explain…(Credit: Facebook screenshot)

For me, one of the most important scientific lessons I ever learned was the vast difference between a correlative and a causal relationship. (Not a casual relationship… this isn’t Facebook.)

A positive causal relationship is when a change in A causes an change in B. If you aren’t a climate denier, a good example is that an increase in greenhouse gases has caused a rise in global temperature.

A positive correlative relationship is when Change A correlates or matches with Change B. This means these two things are positively correlated but aren’t necessarily related to one another.

This is harder to conceptualize off the cuff, so I’ll put it to you this way. From New York to Chicago, large cities have reported increases in violent crime and murder rates during certain periods. Suspiciously enough, sales of ice cream also skyrocket at the exact same time. Egads!

There is a correlation between ice cream and murder!

Another reason to distrust the ice cream man? (Credit: Taylor Dewey, using public commons images 1, 2)

Even though ice cream and homicide rates are positively correlated, we cannot say that they have a causal relationship. In other words,  ice cream consumption does not turn harmless Joe into a murderous monster. One way to test is that  scientists could look at lactose intolerant killers.

But perhaps it is large amounts of money flowing from ice cream sales that is causing the increase. Organized crime thrives, murder rates skyrocket!

Or perhaps little kids (a large demographic of ice cream consumers) are easy targets.

Or perhaps it is just a correlative relationship. If I had to guess, I’d say they are both caused by the hot weather. Good for ice cream, good for irritability and the urge to murder. (But I can’t prove that… I haven’t studied it).

Okay, so you laugh at the ridiculous suggestion that ice cream turns people to murder (the Church of the Flying Spaghetti Monster believes that the lack of swashbuckling pirates has caused global warming.)

Pirates causing global warming? A lesson in correlative relationships

Pastafarians believe lack of pirates is the cause for global warming.

But the problem is that positive correlative relationships do not necessarily lead to causal relationships even when common sense might suggest it. We use cause and effect to make sense of our world. It is easy to point to correlative relationships and say Hey! A causes B!

Its not always that easy though. If we see an increase in both cell phone use and cancer rates, that is a positive correlative relationship.

But does cell phone use cause cancer? This is why scientists spend so much time studying topics. Just because they prove a positive correlative relationship between two things does not mean they have found the cause behind a change.

This is an important realization for anyone who wants to understand science. We can all fall victim to our own common sense and not properly question or test a hypothesis.

So a headline states “High doses of Caffeine linked to Cancer cure in rats!” Ask yourself, how is it linked?

Every correlation you have ever heard of can now be questioned in your own mind. Is there a cause and effect there, or is it just coincidence? How are the two factors really related?

Is it correlation or causation?

Credit: xkcd

Posted on by Kallen in Features, Reporting Science

Add a Comment