PT64.S1.Q22 - microbiologist: because heavy metals

Luluc1234

http://7sage.com/lsat_explanations/lsat-64-section-1-question-22/

I completely understand the argument core and the correct answer. However, upon review, I am confused with the way a Manhattan expert eliminates (A).
https://www.manhattanprep.com/lsat/forums/q22-microbiologist-because-heavy-metals-t5572.html
(See the comment by rinagoldfield. Last post when I last checked. )

Is answer (A) strengthening the correlation between "metal resistance" and "antibacterial resistance" ? According to rinagoldfield, it is. She argued that (A) is incorrect because it strengthens correlation and not the causation. I am confused for two reasons:

1. I think that Mike Kim said somewhere in The Trainer that correlation, although can't prove causation, is perfectly okay to strengthen the causation. (I only glimpsed through the Trainer, let me know if I am daydreaming what this.)

2. I would have interpreted the correlation in (A) as between "not metal resistance" and "not antibacterial resistance" instead of between "metal resistance" and "antibacterial resistance" Am I missing something?

Thank you everyone in advance for trying to help!

Luluc1234

If (A) had said most bacterial "not resistant to heavy-metal" are "not resistant to antibiotics." Then it seem to be more relevant to strengthening the correlation. I would still be hesitant to interpret this as strengthening since it is not strictly applying "no presumed cause, no presumed effect" strategy.

And most bacterial "resistance to heavy-mental" are "resistant to antibiotics" seem hardly a strengthener as well since it is not "presumed cause, presumed effect."

Could people who possess The Trainer kindly help me to clarify this "correlation" strengthening "causation" ? I no longer possess The Trainer.

I am thinking and typing at the same time, sorry if my thoughts are ambiguous and all over the places.

Pacifico

I would help you but I haven't taken 64 yet so I don't want to spoil it. Paging @c.janson35

I can speak somewhat about the relationship between correlation and causation, just not this LR question specifically.

While correlation can never prove causation, it can "strengthen the case for it" by helping to imply that something is the cause of something else, even though there are still other alternatives. In other words, correlation "strengthens the case for" causality by either raising the probability of a certain cause, or reducing the probability of other causes. DNA evidence in crime scenes is a great way to illustrate this. When DNA is analyzed they don't run an entire genome but instead look for a finite number of markers, either six or seven I can't quite remember off the top of my head, but it doesn't matter for the purposes of illustrating my point. So say that the likelihood that someone was a match for all six markers is one in a million. If you ran some DNA from a crime scene through a database of 500 known criminals in the area and one matched all six markers then there is a high probability that they did the crime. However, if you ran it through an FBI database of a million criminals, the probability is higher of hitting more six marker matches, thus you have weakened the case for the original person committing the crime since you have introduced an alternative perp. This effectively weakens the causation, and it makes sense because there are 7,000 people in the world who could match those six markers. Now imagine we decide to test for seven markers and that a match for all seven is one in a billion. If we ran the DNA through that same FBI database, it's highly likely we will only get that one original criminal from the local area. We have used correlation to strengthen causality through correlation. Hope this helps, let me know if you have any other questions.

c.janson35

Haha @Pacifico I was purposely not answering this cause I haven't taken 64 yet either. Can't say I wasn't tempted though

Luluc1234

@Pacifico thank you for getting back! Your explanation and example have helped me to think about this issue more.

@Pacifico said:
If you ran some DNA from a crime scene through a database of 500 known criminals in the area and one matched all six markers then there is a high probability that they did the crime. However, if you ran it through an FBI database of a million criminals, the probability is higher of hitting more six marker matches, thus you have weakened the case for the original person committing the crime since you have introduced an alternative perp.

This seems to me more like weakening by introducing alternative explanations(other people can also match) rather than weakening through correlation, although you did use probabilities and correlation to find the alternative explanation. Let me know if I have misinterpreted your example.

In bare bone formulaic terms, if the conclusion is "A caused B," would you consider the correlation "most As are B"/ "B probably happens when A happens" as strengthening "A caused B"? Your explanation at the beginning helped me to kinda of see how it could do the job. However, I could not think of any LSAT question I have done that does this. Do you have any in mind?

Normally when I see strengthening causal relationship, I go into the default mode of:
1. eliminating alternative cause; or
2. eliminating reversed relationship; or
3. showing presumed cause with presumed effect; or
4. showing no presumed cause, no presumed effect.

Correlation does not seem to fit into any of what I normally look for.

I am not in any rush for solving this issue since it doesn't seem to show up often. And I can always use POE. @Pacifico @c.janson35 However, I would really appreciate if you guys could help me to think a little more about this or perhaps take a look at this Q once you have done PT64.

Luluc1234

@Pacifico @c.janson35 , I got it! Just figured out by answering the same question for another poster. The correct answer is an example of strengthening causal arguments with correlation.

Staying active on the forum is actually helpful!! =D