Sign In | Create an Account | Welcome, . My Account | Logout | Subscribe | Submit News | Customer Service | Tearsheets | Media Kit | Home RSS
 
 
 

Be careful. Correlation does not mean causation

July 11, 2013
By JEREMIE FISH , Adirondack Daily Enterprise

The pressure of our current science culture to publish first and publish fast has its positives and negatives. A major negative is sometimes the published information is wrong, but luckily the system usually quickly weeds out the errors.

However another downside is that in their rush, sometimes researcher's think that they have found the cause of some phenomenon and really they have only found a correlation between the two. Often it is very difficult to figure out if there is truly a cause and effect relationship because there may be some deeper workings at play.

Let me give an example that shows the difference between correlation and causation. Imagine an alien from some other planet who had no idea about our culture visited New York state. Upon visiting the alien noticed that smokers only smoke when they stand outside. The alien then returns to its planet and explains that it has found the cause of smoking in earthlings, whenever they go outside they smoke, so the act of going outside must cause them to smoke.

Clearly going outside does not cause people to smoke, but people do go outside to smoke, mainly because it's the law. So these two events are correlated, but they do not have a cause and effect relationship.

An example of when this has happened recently is the world of statins. Statins are a group of drugs that reduce cholesterol. It was discovered that statins also reduce the risk of heart attacks. So people made that important link, it must be that lowering cholesterol reduces the risk of heart attacks right?

The critical error here was in assuming that a correlation proves causation. It is often more nuanced than that as has been discovered.

It is now understood that many heart attacks happen in people with normal cholesterol levels, and adding a cholesterol reducer does not reduce the risk of heart attack unless it is a statin.

The benefit of statins does not come from its cholesterol lowering effects it comes from something different; some scientists now believe it is the reduction in vascular inflammation that is the root of the cause of reduced heart attacks.

Peeling apart correlations from causations is not as easy as it would seem. That can be seen with statins, yet understandably the urge is there to make the jump from correlation to causation.

So the next time that you see an article telling you that something causes cancer or causes some other disease, be aware that it may be a correlation, and causation does not necessarily follow.

---

Jeremie is a Wilmington resident and a SUNY Plattsburgh graduate. adkscienceguy@yahoo.com

 
 

 

I am looking for:
in:
News, Blogs & Events Web