Q7 - In a study of heart patients

[mshinners]

Question Type:
Inference (Fill-In/Most Strongly Supported)

Stimulus Breakdown:
Heart patient study:
Still waiting to learn if they needed surgery - Less pain
Knew their treatment plan - More pain
Uncertainty (group 1) is more stressful

Answer Anticipation:
That last statement overlaps with the first statement, so the correct answer will likely combine them. The second group is also compared to the first, so the argument can support a relative answer. I'd be looking for something about stress reducing pain.

Correct answer:
(A)

Answer choice analysis:
(A) I would 100% leave this answer on my first pass without selecting it. It definitely matches my prephrase. However, I'd be a little wary of it because it brings up causation (""reduces"") when the stimulus only creates a correlation between pain and uncertainty. That said, after ruling out the others, I'd ultimately be fine with this. Why?
1) It's a Most Strongly Supports question. While correlation doesn't prove causation, it provides evidence for it, so we do have evidence for causation here.
2) This answer is super weak. It's not saying stress always has this effect; just that it sometimes does. It's a lot easier to prove that there was a causal relationship in a few instances of the correlation than to prove that every correlation had a certain causation.
3) The last statements is causal (uncertainty causes stress), so the stimulus does bring up causality.

(B) Out of scope. Beneficial? The stimulus doesn't mention that concept. Even if you equated less stress with beneficial, there's no sign that the pain was related to that state.

(C) This answer choice requires you to assume that stress would affect the severity of the blood flow issues, which there's no evidence to support.

(D) Contradicted/out of scope/degree. While not directly contradicted by the stimulus, knowledge is directly related to stress, so there need not be any relationship between the underlying disease and stress. That said, there is definitely no way to determine what is "probably" the effect and cause.

(E) Unwarranted comparison. The argument doesn't compare the likelihood of needing surgery, just the pain and stress experienced by each group.

Takeaway/Pattern:
Correlation doesn't equal causation. However, correlation is evidence for causation, and in a Most Strongly Supported question with a strong correlation, a causal connection between part of that correlation, and a super weak answer, we can support an answer that argues for causality.

#officialexplanation

[jrnlsn.nelson]

Wait a second, if there's one thing I've learned while studying for the LSAT it's that correlation does NOT mean causation. Now you're telling me that LSAC wants us to affirm logical flaws? That's exactly what (A) is doing, right?

I initially had this circled, and then I thought about it for a second and said "Wait, if I choose this it means that I'm inferring a causation from a mere correlation. There's no way LSAC would want me to do such a thing." So I ultimately chose (C).

Any thoughts?

[emaildanieldrust]

I got this wrong too, but I see exactly why now.

People who were waiting -> less likely to experience pain
People who already knew -> more likely to experience pain

The last sentence tells us to assume that the uncertainty is more stressful than certainty.
So we now know:

People who were waiting -> -likely pain +Stress
People who already knew -> +likely pain -Stress

Diagramming it out it's easier to see why A is right.. the more stress, the less likely to experience pain.

B) Beneficial isn't mentioned in the argument.. but it does make common sense and is attractive
C) The argument doesn't talk about withholding information.. it would take a couple steps to get to this
D) The study was on people with reduced blood flow to the heart. The entire population exhibits this characteristic. We're only talking about the degrees of stress among different types of patients. Furthermore, the argument is silent on the causes of this condition.
E) We're not given any information about which is more likely the case, just information about what is the case if either of two possible states obtain.

The whole argument was really weird and kind of counterintuitive, but perhaps that's what the test writers are checking: to see if we go with what makes the most sense (more stress, more pain) or what the argument actually says.

[jrnlsn.nelson]

Thanks for the reply emaildanieldrust. I follow your reasoning and agree. Yet, I'm still confused as to why (A) is affirming a logical flaw (i.e. a correlation/causation flaw)? Doesn't it seem strange that the answer that LSAC wants us to choose, and to be specific, the answer that "most logically completes the argument," is an explicit logical flaw? I had (A) circled because it does fit the argument, but it's a logical flaw! Why on earth would LSAC want us to endorse this kind of flawed reasoning?

[emaildanieldrust]

I think the only response to that would be the "most" qualifier in the question stem.. meaning it doesn't have to be dead-on-balls accurate (since I'm Vinny Gambini)

That, combined with the fact that the other answer choices bring in information the stim didn't talk about.. I think POE is really the only way to arrive at A being correct

[ohthatpatrick]

I agree with both the anger and the rationalization. Great work.

Question Type: Inference

Task: Read the information given and pick the safest, most provable synthesis.

Tendencies: Almost all Inference questions bring together multiple claims to derive the answer. The connective tissue that lets us bring stuff together is almost always
- Conditional language (if, then, only, unless, requires, all, etc.)
- Causal language (because of this, due to, this allows, that's why, etc.)
- Quantitative language (some, most, all, %, ratio, etc.)

There isn't any conditional or quantitative language going on here, so it's probably testing causality.

Much of causality consists of saying "this group had distinction X. this group exhibited phenomenon Y."

Inference: distinction X may have caused phenomenon Y.

Here, there is a distinction between knowing / not knowing your upcoming treatment and the phenomenon of experiencing / not experiencing pain.

The previous posters are both correct in saying we CANNOT logically prove (A).

Correlations do not prove a causal claim, but, importantly, they do SUPPORT/STRENGTHEN a causal claim.

This question happens to be Inference in the "Fill in the blank" format, but otherwise it would have been a "The statements above, if true, most strongly support which of the following conclusions?"

It could NOT have been asked as a "which MUST BE TRUE" type question.

As such, "most strongly support" / "most logically completes", we don't need perfection. We just need the most supported.

The first responder did a great job of enumerating valid reasons why we have no support for totally out of scope parts of B/C/D/E. Let us know if you find any of them confusing/troubling.

There is nothing in (A) that is totally out of scope. We can't PROVE that stress reduces pain, but we can support it.

A corollary of this happens a lot in Reading Comp inference questions that are also phrased in a "most strongly supported" sort of fuzzy way.

If we read something like
"Gov. Jackson asked farmers to plant more raspberries. This led to a higher than average population of earthworms."

We're allowed to infer (in a most strongly support way)
"If the Gov. had NOT asked farmers to plant more raspberries, the population of earthworms would not have been as large."

We know that this is TECHNICALLY not proven. It's kind of an illegal negation. But it's a supportable judgment. Given that doing X caused Y, we have pretty solid footing to argue that if we hadn't done X, we wouldn't have seen Y (or would have been less likely to see it).

The reason LSAT has us object to correlation/causality arguments is because author's sound CERTAIN of causality in their conclusions. If an author gives you a correlation between X and Y and concludes that "there is reason to believe that X causes Y", LSAT isn't going to have a problem with that. It's the more amped up, "Therefore, X causes Y' that ruffles LSAC's feathers.

Hope this helps.

[jm.kahn]

This question stumped me because I didn't choose A precisely because correlation doesn't mean causation. Based on the credited reasoning, an answer choice that said "It is reasonable to conclude that not experiencing pain sometimes increases the stress one feels", which is that lack of pain is causing stress, than the other way round.

B was particularly attractive because it doesn't infer causation and fits with the argument's evidence. We know from the argument that patients who experience more pain tend to less stressed than patients who experience less pain (who have uncertainty). So some of those who experience more pain are less stressed and therefore pain accompanies some beneficial effects. That is what B says.

Since no choice is provably correct (A commits the logical flaw), why is B not an acceptable choice?

[mshinners]

This question stumped me because I didn't choose A precisely because correlation doesn't mean causation.

I think this comes down to the way that people tend to treat correlation/causation - anything that jumps between them is flawed. If that was the case, we'd essentially never be able to say that one thing causes another, as every causal statement is backed up by volumes of correlative information.

In short, correlation doesn't prove causation. But correlation is evidence for causation, and since this question is asking about what's most strongly supported, and the one premise is causal (not knowing causes stress), we can select that answer.

Note that I'd never pick it in a Must Be True question.

Since no choice is provably correct (A commits the logical flaw), why is B not an acceptable choice?

B brings up something being beneficial, and there's nothing in the stimulus related to that concept.

[RamseyM415]

I had a tough time going with (A) because, although it makes sense given the logic of the argument, it's contrary to commonsense. Curious, do you know of other examples of correct answers that contradict commonsense so explicitly? I can't think of any off the top of my head.

Thanks!!