Wednesday, November 10, 2010

Correlation/Causation

We all know about the fallacy "post hoc ergo propter hoc," wherein it is assumed that because something occurs temporally after another something else, it is thus caused by what came before it. A related mantra in statistics is that "correlation does not imply causation." Nevertheless, confusion reigns here. We see news stories quoting statistics in this precise way almost daily; they tell us we should eat shrimp to avoid prostate cancer, or that walking on the left side of the street will cause symptoms of osteoporosis, or what have you. It's all very silly.

This silliness, however, cannot be so lightly dismissed. It has a dark side. The dark side, I'd posit, is that a lot of people may actually be confused about the differences between correlation and causation. To those of us accustomed to thinking logically, this hypothesis is almost, well, unthinkable. If true, it would provide a solution for the question of how some people can be so incredibly unobservant about their daily lives? Answer: they have the powers of observation but are simply unable to draw accurate connections between one thing and another. Without the power to distinguish between causation and correlation, the world immediately becomes a hazier, more mysterious place.

You must have guessed by now that I have an example in mind--yes, I do, and it's a great, "politically charged" one. Babies. Where do babies come from? It's that perennial mysterious question children ask, which in former times was eventually answered either by experience or by education: sexual intercourse makes babies. It causes babies to happen. Do people really think this anymore, though? I'd hazard a guess that if you carved open the minds of your average joe or jane, the relation between sex and babies would look somewhat vaguer than a good, robust causal link. Obviously, there's a relation there, but causation? Why would sexual intercourse be seen as causal to babies when there are ways to have sex and at the same time make sure you NOT have babies--and there are ways to make babies without involving sex? What's the difference, anyway?

People should apply logic to the situation of their reproductive systems and realize that there's not just a correlation here, but actually a causation. In order to do that, they might first need to break down for themselves what exactly it means for one thing to cause another. They might need to think.

1 comment:

e.k.h. said...

Happy Thanksgiving! Hope you and your family and friends have a blessed one today.