Q9 - Among people who have a history

[ohthatpatrick]

Question Type:
Weaken

Stimulus Breakdown:
Conclusion: Behavior modification is more effective than sleeping pills in helping people fall asleep.
Evidence: People who rely only on behavior modification fall asleep more quickly than do those who rely only on sleeping pills.

Answer Anticipation:
Hmm, pretty tough to argue with this conclusion at first. If behavior modification gets better results than sleeping pills, then isn't the former more effective? If we were going to try to argue that sleeping pills were actually more effective, than how could we explain aways that people who just use behavior modification fall asleep more quickly than people who just use pills? We would have to argue that the two groups of people (behavior mod vs. sleeping pills) had different initial reference points. For example, if I said that people who read a small chapter about English grammar did better on a grammar test than did people who took an 8 month English grammar class, can we infer that the chapter of the book was more effective than the 8 month class? Not necessarily. What if the people who read the chapter were native English speakers who wanted to just brush up on their grammar, while the people who took the 8 month class were foreign language speakers? Even though the chapter-readers did better on the assessment, that doesn't mean that the chapter was more effective than the class, since the two groups had different initial reference points.

Correct Answer:
D

Answer Choice Analysis:
(A) How many hours you sleep in total is irrelevant, since our conclusion is only about effectiveness in "helping people to fall asleep".

(B) Comparing behavior modification people to normal good sleepers is irrelevant, since the conclusion is about comparing behavior mod to sleeping pills.

(C) We don't care whether people have only tried one or have tried both. We just care about what we can / can't infer from the fact that people who only use BM fall asleep more quickly than people who only use SP's.

(D) Yes! This gives us our different initial reference points. If the sleeping pill people are the MOST troubled cases to begin with, than even if sleeping pills are the more effective option, these people might still have considerable trouble getting to sleep. If numbers help, imagine that the sleeping pill people USED to take 2 hours to fall asleep, and now only take 1/2 an hour. That's a 75% reduction in time! If the behavior modification people used to take 1/2 and hour to fall asleep and now only take 15 minutes to fall asleep, that's only a 50% reduction in time. So sleeping pills are more effective, even though the behavior modification people do fall asleep more quickly.

(E) We don't care about the motivation behind the decision to use BM or SP's. We only care about evaluating their effectiveness.

Takeaway/Pattern: The "different initial reference points" is a tough flaw to spot and understand. It's used frequently in anti-causal arguments, when author's are trying to prove that something didn't have an effect or was equally effective to something else. For example, if we say "People who took Patrick's class had the same avg LSAT score as people who studied on their own. Thus, the class was just as effective as studying on one's own." What if the people who took my class started at 140 and got to 160, while the people who studied on their own started at 150 and got to 160. Even though we end at the same point, the different initial reference point shows that taking my class made a bigger difference than studying on their own.

#officialexplanation

[mchuynh]

Can someone explain to me how does D weaken the argument?

Thanks!

[bbirdwell]

The argument says:

Group A uses only X.
Group B uses only Y.
Group B falls asleep faster than Group A.
Therefore Y is more effective than X.

(D) points to the idea that we don't know how bad each group's symptoms are. If it's true, as (D) says, that folks in Group A have worse symptoms, the conclusion might be false. Perhaps using behavior modification doesn't help these people at all, and only works for people whose symptoms are not so bad. This is what is meant by "the most trouble falling asleep."

Remember, a weaken answer does not have to destroy the conclusion, only question it, make it less likely.

[LSAT-Chang]

I correctly chose (D) on this one, but have a question regarding (C). I was really tempted by this answer choice because I thought if many people who use only behavior modification techniques have never used sleeping pills in the past, then I thought this would weaken the conclusion that behavior modification is more effective than sleeping pills since if those people haven't used sleeping pills in the past, how would you know which is more effective than the other?? Does my question make sense? I had two thoughts that MAY make this answer choice incorrect, but just wasn't convincing enough so decided to reach out for help:

1) the word "many" can be ONE, so it doesn't really weaken it so much -- my response to this would be, well.. like everyone says, you don't necessarily have to DESTROY the conclusion, but even weakening it just a LITTLE bit (giving it a little doubt) can still weaken the conclusion.

2) this answer choice may be wrong because the stimulus is concluding from two separate groups (not from people who rely on both techniques, but rather comparing one group who takes sleeping pills from another group who practice behavior modification -- and concludes from THIS that behavior modification is more effective than sleeping pills because those people from group two fell asleep more quickly than group 1) -- so it doesn't matter whether many people who used behavior techniques have never used sleeping pills since after all, the stimulus literally tells us that it is concluding from TWO groups (each of which practice only ONE technique and NOT BOTH).

I guess I sort of answered my own question in the end -- but is #2 the correct reason for eliminating answer choice (C)?

[sch6les]

(1) Behavior modification is correlated with people falling asleep more quickly.
(2) Sleeping pills are correlated with people falling asleep less quickly.
---
(3) Behavior modification causes people to fall asleep more quickly.
(4) Sleeping pills cause people to fall asleep less quickly.

(D) states people taking sleeping pills were before that falling asleep less quickly. Since falling asleep less quickly precedes sleeping pills here, if follows that sleeping pills cannot cause falling asleep less quickly, since a cause must always precede its effect. Therefore, (D) weakens the argument.

Always keep in mind that the LSAT assumes that an effect can have only one cause. If it is stated that sleeping pills cause people to fall asleep less quickly, than NOTHING ELSE but sleeping pills can cause people to fall asleep less quickly.

[dmsqlc1121]

I think the reason that the correct answer is (D) instead of (C) is because the question asks for an answer that "most weakens" the argument. I misread (D) and chose (C) as my answer and it seems arguable that (C) still weakens the argument since it does leaves room for us to question the argument on whether behavior modification is more effective than are sleeping pills. But it most certainly does not "most weaken" with (D) being a stronger answer choice than (C). What do you guys think?

[HGranger]

I may be wrong, but here's why I think C is incorrect:

What I think this choice is trying to insinuate is that if the same people tried both sleeping pills and behavior mod, it's possible that they would find that sleeping pills work more effectively in helping them fall asleep. Basically, "they haven't tried the other option so how would they know what's more effective?"

But the above is not what the answer choice actually says. Rather than saying that there is a Group C that tried both and found sleeping pills to be more effective, it stops short of that and just says "the behavior mod people haven't tried sleeping pills". That statement on its own doesn't tell us anything and doesn't, in itself, cast doubt on the conclusion.

It especially doesn't weaken the conclusion when you consider that there are reasons why sleeping pills may not work for behavior mod people--maybe there is a common allergen in those pills. Maybe they raise blood pressure, and that affects sleep. Just because they haven't tried it, it doesn't mean it could be better.

We can only know that C weakens the argument if we know that a group of people have tried both AND we know the pills were more effective than behavior mod. But we don't know that second half. C as it is requires that assumption to weaken the argument, and we want to steer clear of assumptions dictating logical reasoning, don't we?

[maryadkins]

Yes, good takeaway: steer clear of making assumptions when looking for a weakener.

To run through (C) as well as the other wrong ones:

(A) leaves out behavior modification altogether and so doesn't weaken.

(B) is incorrect because we aren't worried about the rate at which people fall asleep.

(C) is incorrect because it doesn't actually impact whether behavior modification techniques are more effective or not.

(D) is correct (see above)

(E) brings in preferences which doesn't impact the argument.

Hope this helps!

[at9037]

I don’t understand why E is out of scope?

If people are most likely to practice behavior modification because they don’t prefer taking drugs, then the conclusion will be questioned, because it’s not that it’s more effective it’s just that out of the two choices, behavior modification is their only option.

So, my line of thinking was, how can you compare the two, when most of the people have only one option available to them?

[ohthatpatrick]

We don't know WHAT mechanism was involved in "choosing" who was going to be sleeping pill people and who was going to be behavior modification people.

Does it matter whether a doctor chose?
whether an aversion to drugs forced the decision?
whether religious beliefs forced the decision?
whether a random lottery forced the decision?

It kinda does ... to strengthen the argument, we'd like to know that a random lottery forced the decision. If we're doing good science, we want these two groups as equal to each other from the start as we can get.

If we had identical pools of people, and gave pills to one group / behavior modification to the other group, then we'd get a good sense of which treatment was more effective my measuring which group had an easier time falling asleep.

The non-random ways of arriving at our split of people are worse than a random way, but they don't skew the data in any predictable way.

aversion to drugs
vs.
religious beliefs
vs.
geographical limitations
vs.
what your company's health insurance will cover
etc.

Any of these filters could be the reason that some people are doing pills and others are doing behavior mod, and they would all potentially skew our groups in ways that make them less equal to begin with.

So you could say that (E) weakens in the sense that it tells us that a non-random filter is influencing how the groups were divided.

But (D) is a much better answer because it ALSO tells us that a non-random filter is influencing how the groups were divided, and (D)'s filter has a logical connection to what we're studying.

(E) might skew the data by creating two different groups that are unfair to compare.
(D) definitely skews the data by creating two different groups that are unfair to compare.

Does that make sense?