Does causation equal correlation?

Comments

• 1 Like

  BinghamtonDave Alum Member ⭐ 🍌🍌

  February 2017 8716 karma

  I wouldn't say that causation "equals" correlation, because they are two separate and different concepts, instead I would rephrase that by saying that causation implies correlation. If you have a causal phenomena, it follows that there is a correlative one. In an analogous vein if you have all of something then of course you have some of something.
  Consider an example:

  The economic failures of Communism caused the collapse of The Soviet Union.

  Implicit in that statement is the idea that those two concepts are correlated. Because if they weren't in any way, we couldn't arise to the standard of that statement. For further explanation of this, please refer to
  https://7sage.com/lesson/causation-theory/

  I hope this helps
  David
• CinnamonTea Member

  February 2017 550 karma

  Thanks!
• alex.e92 Alum Member

  February 2017 239 karma

  Basically what @BinghamtonDave said. Correct me if i'm wrong, 7sage, but I can't think of any example that would counter this: Correlation (in either direction) is necessary but not sufficient for causality. So, if I notice a correlation, or a relationship between two things, that by itself isn't enough to prove that one of those things is causing the other thing to occur. On the other hand, if there is causation between two things (we know that one is causing the other), a relationship between those two things (correlation) is a given.
• 1 Like

  Cant Get Right Yearly + Live Member Sage ⭐ 🍌 7Sage Tutor

  February 2017 27901 karma

  @"alex.e92" said:
  Basically what @BinghamtonDave said.

  https://media1.giphy.com/media/FoauNqTIVwuIg/200.gif#1
• 1 Like

  BinghamtonDave Alum Member ⭐ 🍌🍌

  February 2017 8716 karma

  @"Cant Get Right" not embarrassed to admit, my #2 all time favorite movie.
• Q.E.D Alum Member

  February 2017 556 karma

  On the LSAT, yes.

  Philosophically, no. Actually, I remember an RC passage about discoveries in Chaos Theory that implied there may be some physical processes we'll never understand because they behave chaotically, meaning experimental results are so sensitive to initial conditions - perhaps arbitrarily so - as to be unrepeatable. In a nutshell, there could be a causal relationship between A and B without any discernible correlation.

  A more mundane example, though not really a counterpoint, is that experimental or observational conditions aren't controlled enough to account for disruptive factors, which could conceal otherwise observable correlations between causally related factors. For example, we could do a study of smokers and non-smokers to show no correlation with heart disease by choosing an older sample of nonsmokers than smokers. But, as I said, that's not really a counterpoint, since in that case you choose just the "biased samples" flaw or whatever.
• Aidoe Free Trial Member

  February 2017 edited February 2017 236 karma

  @Q.E.D Sounds like your first point is the same as the second. Just because in the first case the correlation is not discernible, doesn't mean there is no correlation. I always thought that by definition, causation necessarily implies correlation. A correlation is just a mutual relationship among independent variables, and causation gives order to that relationship.
• 1 Like

  Q.E.D Alum Member

  February 2017 556 karma

  Hey @Aidoe
  If you take correlation in the statistical sense of two factors having an improbable (assuming randomness) amount of covariance, then they are quite different. There would be absolutely no statistically significant relationship between A and B in a chaotic system, even though there are deterministic laws governing B based on A. It's that system itself that is chaotic. Even accounting for all other factors, there could be absolutely no statistical relationship there.

  That's different from the second point, where the lack of a discernible correlation is simply due to ignorance of other influences.

  Chaos Theory is often misinterpreted as being based on ignorance of all the factors, so to speak, which is way off base. It's a deep, deep discovery about the inscrutability of some non-linear systems. The implications for our ability to understand the natural world are profound.

  But if you take correlation to mean just a non-random relationship (not just improbable), one that is to some extent determined, whether observed or unobserved, then you're right, causation does always imply correlation. But by that definition, it's hard to see how correlation doesn't imply causation, which in the leanest scientific sense just is such a deterministic regularity the world.

  So as long as we're distinguishing the two, I would just point out the exception for chaotic systems.

  Not really relevant for LR questions, though. Just pointing that out because CT is quite earth-shattering precisely because it undermines that assumption about the world.
• 1 Like

  Aidoe Free Trial Member

  February 2017 236 karma

  @Q.E.D Oh God, I asked for that beatdown but it was worth it. Interesting stuff.
• 1 Like

  Q.E.D Alum Member

  February 2017 556 karma

  Lol @Aidoe I felt like you asked a poignant and challenging question, very much to the point. It made me feel like a bloviating asshole for throwing things out there and expecting them to be uncritically accepted.
• Aidoe Free Trial Member

  February 2017 236 karma

  No, you know what you're talking about and I certainly appreciate you getting into the deeper philosophical stuff even if it's not necessary for the LSAT.