Causation vs. Correlation vs. Conditionality

monica123

Hi All,

Does the below statement make sense? I am trying to get a big picture view of these 3 ideas.

CAUSATION --> PERFECT CORRELATION* < ---> CONDITIONALITY

Causation implies Perfect Positive Correlation.
Correlation (regardless of strength) does not imply causation.

Causation implies Conditionality. I.e. If A causes B, we can also say A is sufficient for .
Conditionality does not imply Causation. Just because A is sufficient for B, does not mean that A causes B. It could be, for example, that A

Perfect Correlation (and only perfect correlation) implies conditionality
Conditionality implies perfect correlation.

*I am assuming A ALWAYS causes B. Is this the case on the LSAT? Can we say that if A causes B, A always causes B?

Accounts Playable

No, the relationship you have isn't true. Conditional reasoning and causal reasoning are different; sometimes they can be mutually exclusive and sometimes they can't be mutually exclusive.

A perfect correlation is something in which there is a perfect positive correlation or a perfect negative correlation. Say we have this relationship: every single time I see a movie in a movie theater, it's night time outside. In other words, there is a perfect positive correlation between me going to a movie and it being dark outside (correlation is 1). This doesn't imply that me going to a movie causes it to be dark outside or that it being dark outside causes me to go to a movie. It's just a coincidence that this is the case. In reality, the true cause of my chosen movie times is that the afternoon is the only time that all my friends can go with me.

It is true that causation does imply some type of correlation, it just doesn't have to be perfect. There is a causal relationship between smoking and getting lung cancer, but there are people who smoke who never get lung cancer. For this same reason, causation doesn't imply conditionality.

The bi-conditional relationship between perfect correlation and conditionality gets a little into the weeds about what it actually means for something to be necessary. Say I have a data set of every single time I've ever gone to a movie. Based off what I stated earlier, we know that each of those times was at night. But, what about the next time I go to a movie? Does that imply it will be at night? Not necessarily! Things could change! Maybe I decide to go alone in the morning; maybe my friends' schedule changes so that they can go earlier.

On the other hand, say I have a completely different data set (imagine two columns in Microsoft Excel). In one column is the name of every single United States president. The second column lists that president's corresponding nation of citizenship. Every single president has been a United States citizen. It's also the case, by necessity (the Constitution), that every single future president will be a United States citizen. So in this scenario, it's reasonable to assert a conditional relationship: if someone is a United States president, then that person is also a citizen of the United States.

What about the other way? Pragmatically, we know that being a United States citizen doesn't imply being a United States president. For practical purposed on the LSAT, your intuition will help you here. In reality, we'd need data points that suggest this relationship.

monica123

@"Accounts Playable"

Thank you! I now realize this relationship is completely context dependent. Your explanation is amazing!