Correlation versus Causation, part 1

Correlation is not causation!

Obvious, right? Is some contexts, yes it is. But the issues get complicated and in fact one of what I consider the “Big Three” LSAT flaws is the Correlation/Causation flaw. To do well on the Logical Reasoning of the LSAT (which as of July 2024 will constitute two thirds of your LSAT score, not that logic games have gone by the wayside) you need to recognize this flaw and know what to do with it.

I’ve posted a video about this flaw, in which I go into quite a bit more detail than I’ll include here; you can watch it here. (I can’t embed it on this page because the file is too large.) In the video I discuss:

  • The Big Three flaws:
    • Correlation/Causation;
    • Necessary/Sufficient
    • You haven’t proved you’re right so you’re wrong
  • What issues you don’t have to worry about:
    • Random noise
    • Sample size (rarely)
    • Biased collectors (though that’s a huge issue in reality)
  • What issues you have to worry about occasionally:
    • Timing
    • Biased or otherwise unrepresentative sample

I then focus on the most common issue in correlation/causation arguments, which is the possibility that another, as-yet-undiscussed factor is the actual cause.

These issues can arise in any conclusion-based argument but or most frequently found in weaken, strengthen, necessary assumption, and flaw questions.

Part two of this series is discussed here.