Correlation Question

[voodoochild]

Experts,

When I say : People who wear white shirts are more likely to eat mangoes. Is this correlation?

If yes, is there a way we can "imply" causality using the phrase "more likely"? (I know that by saying X is more likely to cause Y ==> clearly implies a causality because we ahve used an explicit causal indicator)

I am looking for a context in which by using "more likely" I could imply causality without using "explicit" causal indicator.

Any thoughts?

[Nikki Siclunov]

"More likely" in and of itself does not present a causal relationship: it is an indicator for a correlation between two independent variables. Indeed, in the example you provide, wearing a white shirt correlates with eating mangoes. Causation is an argument function: it is usually something we conclude from a given set of facts, correlations, comparisons, etc. So, if you said something like, "Therefore, by wearing a white shirt tomorrow I am more likely to eat mangoes" - this conclusion would imply some form of causation between the two.

[voodoochild]

Nikki,
Thanks for a detailed response. After reading your response, I am a bit confused about my discussion with Adam:

98% of mothers who didn't receive the daily dose of adult nutrition during the pre-pregnancy program gave birth to babies with lower immunity. Hence, if a mother forgot to eat proper nutrition during the pre-pregnancy program, there is a good chance that the babies will be born with lower immunity than if mother didn't forget to take the daily dose of adult nutrition during the pre-pregnancy program.

thread:http://forum.powerscore.com/lsat/viewto ... hers#p5272

Do you think that the above argument is implicitly causal? I am a bit confused.

Thanks

[Steve Stein]

Hi Voodoo,

In response to your question, if your example were to show up as a causal argument on an actual LSAT, it would be worded more along these lines: "...Hence, mothers who did not receive proper nutrition during their pregnancies must have, as a result, increased their chances of having babies with low immunity. "

The test makers go out of their way to make causal arguments fairly explicit, as Nikki pointed out earlier in this thread.

I hope that's helpful!

~Steve

[Adam Tyson]

I'll add this to the discussion - I don't believe the LSAT has any implicit causal arguments. When you see a correllation, don't make any causal assumptions - if you do, you're making a classic Flaw in the Reasoning ("voodoo assumed, without justification, that one thing caused another based solely on evidence that the two things are correllated.") Correllations are only correllations - it's not until someone draws a causal conclusion that it becomes a causal argument. That's what Nikki meant about causation being an argument function - someone has to explicitly make that argument.

Your initial example on this thread is not causal, but conditional - "people who" being the key indicator of a sufficient condition. If the reasoning is conditional, use your conditional reasoning techniques to analyze it, and don't try to make it causal by making an unjustified assumption.

Adam

[voodoochild]

Adam,
Thanks for your reply. I am not sure about when you say 'the author will explicitly' mention causality. Do you mean that the author will explicitly use "causal indicators" mentioned in your book? There are a lot of examples that don't use explicit causal framework.

Let's take Nikki's example ("Therefore, by wearing a white shirt tomorrow I am more likely to eat mangoes")

OR
we could also consider : PT40 - S1 #6 (I found this from another parallel thread :http://forum.powerscore.com/lsat/viewto ... =12&t=2381 - you are there on that thread)


In official LSAT question, the author doesn't use causal indicators to say that, because diets were different, frequency of dental problems was different. In fact, the conclusion says "Most likely, ...." From our discussion, "most likely" is a correlation indicator. Do I have to assume that the conclusion is correlation based? I don't think so. In my opinion, if the premise is around Correlation, there is a good chance that the conclusion will be causal -- implicit or explicit.

For instance,
Arg#1:
Premise1: Whenever I get up early in the morning, I solve all assumption questions with 100% accuracy
Conclusion1 - Most likely, getting up early is better than sleeping late.

Arg#2:
Premise = Premise1
Conclusion2 - Most likely, getting up early helps Voodoo in solving Assumption questions.

Arg3:
Premise = Premise1
Conclusion3 - Most likely, the brain must be getting prepared for solving tough problems.


We can see that Conclusion1 is not causal. Conclusion2 and Conclusion 3 are. I have used the same premise to come up with three different causal and a non-causal arguments. I think that the meaning of the argument is the key. Also, Argumetn#3 doesn't use any 'causal indicator' such as 'helps' in argument2. Conclusion3 just gives a reason why I am able to solve tough problems when I get up early. This argument is very similar to PT40 - S1 #6 [This paragraph is my opinion. I am not sure. Please correct me if I am on the wrong track.]

Please let me know your thoughts. I would appreciate your opinion.

Thanks

[Adam Tyson]

I'm not saying that the author will necessarily use any of the causal indicators we list in our materials. Those are common indicators, but it's hardly an all-inclusive list. I'm saying that causality is not implied - if you infer it based solely on a stated correlation, then you're making the same sort of error that we jump on when LSAT authors do it. The author has to not only set out the correlation, he has to then make a conclusion based on that correlation, and that conclusion has to indicate causality in some way.

The effect of "more likely" will vary depending on the structure of the argument. It could indicate causality, or it could continue a conditional relationship. I don't think we can apply a blanket rule to that phrase to say that's it always going to be one thing or the other.

Adam

[voodoochild]

Thanks Adam for your detailed response. Two quick questions:

(1) do you mind commenting on my three arguments? (I spent a fair amount of time thinking and re-thinking such situations )
(2) You have mentioned that "more likely" could be used in "conditional set-up" -- do you mind giving an example? That would be really helpful.


thanks
Voodoo Child