How to identify correlation vs causation?

[awesomefun2010]

I have a question on wording.

If a statement says something along the lines of: Physics students are more likely to do well on the LSAT than Chemistry Students.

Does this indicate correlation or causation? Specifically, what does the phrase "more likely" mean on the LSAT in such contexts.

Another question: As "x" increases, y increases. --> does this indicate correlation or causation?

I have been so confused on this. Thank you so much

[Robert Carroll]

awesome,

Your statement about physics students would indicate that "being a physics student" and "doing well on the LSAT" are correlated, or at least correlated better than "being a chemistry student" and "doing well on the LSAT". There is no causal language in there. LSAT stimuli often contains flaws where a premise like the statement you gave would be the basis for a conclusion implicitly or explicitly about causation.

For example: "Physics students are more likely to do well on the LSAT than chemistry students. So if you want to increase your LSAT score, you should change your major from chemistry to physics."

The conclusion is the second sentence, and I wonder "Why should I change majors?" This author is thinking that the physics students do better BECAUSE they are physics students, so that being a physics student causes the better LSAT performance. But the first sentence, which is identical to your example, is merely a claim about correlation. So this author thinks that correlation also indicates causation, which is flawed.

My sample (flawed) argument there uses implicit causation. The only reason to even think that switching majors will help the LSAT score is a mistaken assumption that the correlation in the first sentence is also causal.

Your second example also looks like correlation to me. It doesn't claim that an increase in x will produce an increase in y, but just says that as one increases, there is an increase in the other. Without more information, that just sounds like an observation of a correlation between the two.

It's easy to think that such a relation between the two indicates some kind of causation, but that's a mistake, and a mistake that many stimuli commit. It is, in fact, the whole point of those bad arguments that they make that leap. So, when a statement like your examples is made without further info, be on the lookout for erroneous claims about causation.

Robert Carroll

[awesomefun2010]

Thank you!