Correlation vs. Causation: Differences Between Likely and Direct Causes

Increased cell phone use leads to autism, at least according to genengnews.com, which features Tamara Mariea's correlated research on cell phones and autism in America.

The initial reaction would most likely be: "OMG, I gotta quit using cell phones and sharpen my pencils and go back to snail mail." Relax!! It's correlated research.

The most important thing to remember is that correlation and causation are not the same thing. Correlation, according to dictionary.com, is "a statistic representing how closely two variables co-vary." In other words, when one variable is studied next to another, an effect is likely to happen. A causation is when two variables directly affect each other. For example, if you put your dog outside at night and it gets sick, and this happens many times, it's likely that your dog gets sick because it is put outside. It might not be the cause, however. A virus or bacteria might be the cause that results in your dog being sick.

It is essential to distinguish between the two. Correlation asks the question: What relationship exists between the two variables? What connects or separate them from each other?. Experimental research put the emphasis on what would happen if one variable is changed whereas correlated research does not try to influence the components of an experiments but rather to observe the outcome of the two events and offer statistical data as proof.

Could you distinguish what is causal and what is correlated?

-Bullying harms kids' mental health.

-Stress of watching the Super Bowl can be hazardous to heart.

-TV raises blood pressures in obese kids.

-Deep voiced men have more kids.

-Women who don't use condoms feel happier (I'd like to be in this research).

-Surgeons with video game skills perform better in simulated surgery.

If you think even one of the above hypothetical researches is causal, you are so wrong. According to Jon Mueller, the media often attaches false title to correlation research to attract viewers. Researches that seem to suggest events, rather than support them, are less likely to bring large volumes of audience. After all, who cares about the correlation that listening to music while studying seems to suggest bad grades; people only care that listening to music causes bad grades. Therefore, the next time you hear some generic statistics that seem interesting, remember to do some more research and hope that it's a causal experiment; otherwise it wouldn't be as significant as the media has made it out to be.

Knowledge is not power; the correct knowledge is.