Q8 - Doctor: In three separate studies, researchers

[bbirdwell]

Once again, we must first closely consider the logic of the original argument.

It has to do with CAUSE. Notice the conditional nature of the conclusion, also. "IF night-lights CAUSE nearsightedness, the effect disappears with age."

This is based on the evidence that:
1. Studies compare children who sleep w/nightlights to those who don't
2. In first study, children w/nightlight = more likely nearsighted
3. Next two studies, no correlation.
4. Children in first study = younger than those in other two

Focus on weakening the conclusion when you consider answers.

(A) This answer merely fails to satisfy the condition presented in the original conclusion: "IF nightlights cause nearsightedness." It does not damage the conclusion itself.

(B) irrelevant.

(C) irrelevant.

(D) directly weakens the argument by essentially disqualifying 2 of the 3 studies on which the conclusion is based. Throw out those two and all you're left with is the original, where a correlation was found.

(E) is close, but ultimately doesn't really weaken the conclusion that IF the nightlight CAUSES the nearsightedness, it disappears with age. Why? Because we don't know what caused the nearsightedness of the "several children" cited.


#officialexplanation

[expfcwintergreen12]

What does the doctor conclude? Does he conclude
If A causes B, then C.
Therefore C.

If "therefore C" is actually his conclusion, then I offer my humble apologies for wasting everyone's time. If "therefore C" is actually the conclusion, then I grant bbirdwell's explanation, cogent and beauteous. The doctor's argument assumes the antecendent "A causes B" of the conditional. By falsifying it with answer choice D, ("A does not cause B") his conclusion does not follow.
To me, however, his conclusion is not "therefore C" ("Therefore, the effect disappears with age." But rather his conclusion is
"If A causes B, then C." His conclusion is actually made up of an antecedent and a consequent. His conclusion does not affirm the consequent.
If I am correct about his conclusion being conditional, then as the many truth tables I slaved over during logic 101 informed me, his conclusion is not weakened by falsifying the antecendent, in fact, falsifying the antecendent makes the conditional statement true, necessarily true, damn true, truey true.
And if I am right about the conclusion being conditional, and about answer choice D attacking its antecedent, then, per the laws of logic, D cannot weaken the conclusion.

[bbirdwell]

Okay. First of all, I think this is a really tough problem in some ways -- unusually so for a #8, actually. Second, you're absolutely right that the conclusion is conditional. And while I don't speak the language of "antecedent and consequent," and I wouldn't know a logic table from a ping-pong table, I do get perfect scores when I take practice LSATs, and near-perfect when I take the real thing. Here's what I have to offer:

It's an argument based on three studies. Answer choice (D) says two of the studies were bunk. That weakens it, plain and simple. This is a useful technique when dealing with arguments and answer choices that are clearly designed to be complicated -- simplify them rather than playing the game that the test-writers want you to waste your time playing.

Process of elimination is also a useful tool here, I think. To me, D and E are the only real candidates. And if you can see why E is wrong, you're left with D.

D, by disqualifying two of the three studies, does more than attack the "antecedent," I believe. If those two studies cannot even meet the criteria of drawing a causal relationship in the first place, then we can't use them to support our conclusion, period. Right? This is because our conclusion is based on that causal relationship. Finally, if those studies cannot satisfy the condition, and therefore cannot be used, then we have no "older" children to compare with those children from the first study. Thus, we no longer have any evidence at all to suggest that the effect "disappears with age." We simply have one study that says such an effect is likely to exist.

Is the conclusion proved false by D? Nope. Does it need to be? Nope. Is it made less certain, considering the evidence on which it is based? Yep.

[ManhattanPrepLSAT1]

There's one point that I'd like to mention and this may be obvious and so that's why it wasn't mentioned. In my mind, the most important word in answer choice (D) is the word "any."

We're asked to weaken the doctor's argument. The doctor's conclusion is that "if night-lights cause nearsightedness, the effect disappears with age." That is a conclusion regarding a causal relationship. Answer choice (D) does not undermine the causal relationship between night-lights and nearsightedness but it does undermine the conclusion regarding that causal relationship, "that the effect disappears with age."

Does that help clear this one up at all? I completely agree with Brian that this one is ridiculously tough for being so early in the section! I've struggled teaching this one in class for the past several years - it's not easy to see!

[cyruswhittaker]

So let me make sure I understand..

D is correct because it attacks the assumption that the conclusion is based on: that underlying the two surveys, there was evidence of a causal relationship between night-lights and nearsightedness.

If this were not true, then essentially the studies would prove irrelevant to the conclusion. And choice D attacks this assumption, by saying that no causal relationship could be asserted from the studies due to their inherent flaws.

And E is incorrect because although on the surface it appears to present a counter to the conclusion, it really does not since we cannot be sure that "several of the children" actually were near-sighted as result of the night-lights. If this were not the case, then it would be irrelevant to the conclusion since the conclusion has the condition "If night-lights..."

Are these explanations correct? I had a challenging time figuring this question out, and am still a little uneasy about it, but just want to make sure I have the right idea. Thanks.

[ManhattanPrepLSAT1]

Not exactly.

Answer choice (D) says that the studies could not provide support for any conclusion regarding a causal relationship.

This is not the same thing as saying whether there is a causal relationship. So in this case the conclusion regarding a causal relationship is whether the effect disappears with age.

The problem with answer choice (E) is that "several" is not enough to tell us whether the effect disappears with age. We have nothing to compare to from the stimulus. All we were give above is the term "more likely."

Hope that helps put some perspective on this one...

[sissixz]

Well, sorry, still seems too difficult, and confused to me.

I can understand you guys are trying to prove that "D" points out no causal relationship between nearsightedness and night-lights, so it weakens.

But the author's conclusion is a conditional one, he simply hopes if there is chance to have some causal relationship, that it will has something to do with ages.

When I read about this question, I was searching for an answer about negating relationship between age and nearsightedness. So E feels right for me, though not that sound for the word "several", since I thought some sort of weakening still counts.

Hope someone can put more insights.

It's unreasonably hard on #8!!!!!!!!!!!!!

[noah]

But the author's conclusion is a conditional one, he simply hopes if there is chance to have some causal relationship, that it will has something to do with ages.

And, what (D) tells us is that we don't have any reason to think it has something to do with age. The evidence that led him to think that has now been shown to be flawed.

Analogously: I conclude that if owls do sing, they only do so at night. Why? Because owls were observed 3 times, in the morning, evening and night. The scientists observed that the birds made song-like noises only at night.

(D) it turns out that the observations done in the morning and afternoon were not done correctly - those were hawks, not owls!

(this is not particularly LSAT-like, in that it contradicts a premise, but it should get the point across)

When I read about this question, I was searching for an answer about negating relationship between age and nearsightedness.

And what is that connection based upon? The fact that studies of older kids didn't show a relationship while the one of younger kids did. So, if we invalidate the studies of older kids, all we have is one study showing a connection for young kids - nothing to draw a conclusion about how that compares to the effect on older kids.

Make sense?

[joseph.m.kirby]

I mistakenly chose (E) on this question. I now understand why it is wrong.

The stimulus notes that the later studies found no correlation between night-lights and nearsightedness.

(E) says that several of the children who had slept with night-lights as infants were nearsighted; however, (E) does not say that there is a correlation. Thus, (E) does not weaken the argument. Perhaps several of the children, in other studies, who slept with night-lights as infants were nearsighted as well....

Overall, to weaken this question, we are looking for an answer that attacks the correlation, (D).

[ManhattanPrepLSAT1]

Exactly right Joseph!

Answer choice (E) is simply too weak to suggest that the children who had slept with night lights and were still nearsighted were nearsighted as a result of having slept with a night light. Since the conclusion regards causality, we need to weaken the conclusion regarding why the children developed nearsightedness. Answer choice (D) addresses the issue of causality directly, while answer choice (E) leaves open the question of why the children developed nearsightedness.

Hope that helps!

[samjcg]

Here is my thoughts.

A) The study mentioned in this answer is comparing wrong subjects. They are comparing INFANTS to INFANTS. We need a study that compares INFANTS to CHILDREN. Irrelevant.

B) This is irrelevant. On Average? Ok, so what?

C) Study fail again. Infants did not sleep with night-lights on but children had. Wrong direction. Irrelevant.

D) Correct. This is a sample error. The premise that was supporting the conclusion did not examine enough children to provide significant support for their claim. This weakens the argument.

E) "several" is too weak. Also this study was based on children who were OLDER THAN THOSE IN ANY OF THE FIRST THREE STUDIES. Moving away from the argument. Irrelevant.

[ohthatpatrick]

Nice explanation. Let me add a few thoughts.

Weaken

Conc:
If night-lights cause nearsightedness, the effect disappears over time.

(why?)
Prem:
3 studies involving infants/nightlights:
- the one involving the youngest infants found a correlation between sleeping with nightlight and nearsightedness
- the two studies involving the older infants did not find a correlation

How do we weaken this?
When you're weakening a conditional claim, such as this conclusion, you can only disagree with the second half, which in this case is that the nearsightedness disappears with age.

The author isn't actually claiming that nightlights cause nearsightedness. He's saying that IF they do, the effect disappears (trusting those 2 studies that showed no correlation).

We need an answer that would help us go against those 2 studies.

(A) this strengthens those 2 studies ... it also finds no correlation

(B) this starts with children who are ALREADY nearsighted, so it's useless for assessing whether nightlights cause nearsightedness

(C) this seems to suggest that nightlights for infants as well as for older children did not cause nearsightedness ... again that goes with the 2 studies

(D) this attacks the credibility / sample size of the 2 studies

(E) this involves older children, making it suspect ... it also doesn't give us any correlation .. saying that "several" kids were nearsighted doesn't help us judge whether there's a real connection.

I find it very challenging when Weaken/Strengthen questions attack the credibility / trustworthiness of the evidence. It rarely happens, but it's important to keep in the back of our minds the possibility that an answer can function that way.

Hope this helps.

[csunnerberg13]

I'm confused about when an answer choice can attack the credibility of evidence and still be the correct answer. It seems like D should be ruled out because it's undermining what we are already told...why is it okay here to choose an answer choice here that basically just says that premise isn't true? And when is it NOT okay to do so?

[ohthatpatrick]

In short, it's ALWAYS fair game to attack the credibility of the premise.

However, in practice, it normally only comes up in Strengthen/Weaken (and only in 5-10% of those at most).

Be careful, though, we are not FALSIFYING or CONTRADICTING a premise.

We are saying that a premise is TRUE, just not worth much.

Consider this argument:
Stephen Colbert usually has liberal minded guests on his show and, when asked by Colbert about their feelings towards George W. Bush, almost all of them have said that "George W. Bush was a great president". So, clearly, liberals recognize that conservative President's virtues.

Which of the following, if true, would weaken?
(A) The question posed to the guests was, "George Bush: A great president or the greatest president?"

(This is borrowing from real life, because Colbert used to do this)

This answer Weakens the argument. The premise is still true: it's still accurate to say that the vast majority of liberal guests said "GW was a great president".

But our impression of the QUALITY of that evidence has changed. We think, "OH, so it was a biased question, and the vast majority was picking the LESS flattering answer."

This is how it's legal to undermine evidence.

This mode of strengthening/weakening generally only comes into play when the evidence used is:
survey
study
sample
experiment
comparison/analogy
testimonial

You can truthfully report on any of these things, but then address their credibility by calling into question:
biased questions
truthful answers
large enough sample
representative sample
sensible methodology
fair comparison
relevant expertise of speaker
potential bias of speaker

[redcobra21]

Hey guys

I've looked over the responses but still have some confusion.

(1) First, I recall that the Manhattan strategy guides mentioned that if you are looking to weaken a conditional, the best method of doing so is to attack the NECESSARY portion of the statement. When I read this question, it looked like the LSAT was testing this concept. (E) seems to directly attack the NECESSARY part by saying that the effect actually does NOT disappear with age. After all, the fourth study showed that there were older children where the effect had not disappeared (i.e. they were still nearsighted). On the other hand, (D) seemed to be attacking the SUFFICIENT portion of the statement. Matt brought up how the answer needs to address the issue of causality, but I don't really see how that is relevant since the doctor is positing IF this is the case (in other words, even if you did say that there was no support for a causal conclusion, the doctor could say that does not matter because he was only talking about what would happen IF there was a causal connection). Am I missing something here? I've applied the book's strategy in other previous PTs and got the correct answer, but what everyone has been saying on this thread appears to go against this trend.

(2) Second, I had a question about the critique of methodology described in (D). The answer literally just says that the two studies did not "examine enough children to provide support for any causal relationship." But without further elaboration, how can you accept this statement at its face value? Statistically representative samples could quantiatively have extremely low absolute N-values, and just because the answer choice says that there were "not enough" does not mean that this is not statistically significant. Is a statement like this enough to assume that the two studies should be totally discredited as Matt and Birdwell suggest?

Coupled with the attractiveness of (E) and the problems with (D), I used Manhattan's method of process of elimination, picked (E), and moved on.

Am I missing something here? Sorry to beat a dead horse. Thanks in advance for your help

[amil91]

Perhaps I am misunderstanding the conclusion, but I understood 'the effect disappears with age' as meaning that an older child sleeping with a night-light is less likely than a younger child sleeping with a night-light to be nearsighted; not that a child who as an infant slept with a night-light was nearsighted, but 'grew out' of that condition with age. So with this in mine I eliminated E because it does not address the issue of older children sleeping with night-lights, only older children who had slep with night-lights as infants. I unfortunately picked A, but I now see why that is incorrect.

[ohthatpatrick]

Good follow-up questions.

I'm not 100% sure how we should interpret the conclusion; I'm sympathetic to the alternative interpretation just suggested.

However, for me, saying that the effect disappears with age sounded more like the nearsightedness (the effect) wears off.

The alternative interpretation just offered sounds more like night-lights are a CAUSE early on, but are no longer a CAUSE with increasing age.

Luckily, I don't think we need to be sure of either interpretation to get rid of (E).

The previous poster who was saying, "Don't we attack a conditional by attacking the necessary idea?" is correct. We do.

Does (E) suggest that nearsightedness did NOT disappear with age?

Not really. It says that "several" of the 100 children were nearsighted. That is numerically very weak, and there's no way to know if that's a statistically significant quantity. The previous studies showed "positive correlations" between night-lights and nearsightedness. Correlations are suggestive of causality. This is just saying several of the kids were nearsighted. If I said several of the kids were left handed, would we speculate that the night-lights had caused them to be left handed?

We would need language in (E) that sounds like there is a positive connection between night-lights and nearsightedness to push back against the necessary idea in the conclusion. Otherwise, if we're saying 97% of the night-light kids in this study were NOT nearsighted, it seems to SUPPORT the conclusion ... the effect does not seem to be manifested for kids of this age.

When an argument deals with correlations, there is always room for exceptions. Just because SOME students get A's without studying doesn't undermine the idea that there is a positive correlation between studying and getting A's. So be wary of weakly worded answer choices.

Another question asked by the previous poster was:
The answer literally just says that the two studies did not "examine enough children to provide support for any causal relationship." But without further elaboration, how can you accept this statement at its face value?

We're not supposed to attack the credibility of answer choices on Strengthen/Weaken.

We can accept it at face value because the question stem reads, "Which of the following, if true ..."

Hope this helps.

[arash.nouraee]

Just wondering, if (E) said something like: "in the fourth study with 100 kids, MOST of them who used night lights as infants were nearsighted," would that weaken?

My reasoning is that the argument says the effects wear with age, so if when they got older most of the kids who used night lights were nearsighted, then it would show that the argument is invalid.

I just want to get a sense of how game-changing the word several was in this argument.

[ohthatpatrick]

I think you’ve nailed the problem with (E).

To push back against the author, we’d need to show that night-lights DO still have an effect on older children.

Does (E) count as a piece of evidence towards making that argument?

No. Let’s say that of those 100 kids, 80 of them slept with night-lights when they were children. 3 of them (several) were nearsighted.

Would that statistic strengthen the claim that night-lights cause nearsightedness? No.

But, as you said, if we knew that MOST of the 80 kids who had night-lights also were nearsighted, THAT would actually be a positive correlation, which WOULD strengthen the idea that having night-lights might have some causal effect on nearsightedness.

[phoebster21]

Exactly right Joseph!

Answer choice (E) is simply too weak to suggest that the children who had slept with night lights and were still nearsighted were nearsighted as a result of having slept with a night light. Since the conclusion regards causality, we need to weaken the conclusion regarding why the children developed nearsightedness. Answer choice (D) addresses the issue of causality directly, while answer choice (E) leaves open the question of why the children developed nearsightedness.

Hope that helps!

I'm sorry, but I don't quite understand why the issue of "whether it is or isn't causal" is important here. I thought the issue is whether or not you can draw any specific conclusions about the nature of this alleged causal relationship, i.e. that this effect decreases with age.

So whether or not answer choice E establishes night-lights as causing the nearsightedness shouldn't weaken it right? In order to draw a conclusion like this one about (possible) causal relationship, you'd have to have multiple points.

It's not possible to infer an increase/decrease without having at least 2 points (so you can calculate a difference between them). So here, the correct answer is saying that we only have 1 point (along a continuum that includes 3 points) so we can't draw any conclusion about a decrease.
Its like saying, you can't tell if something is increasing in speed just from clocking them at one point. You need to clock them once, and then clock them 10 seconds later, and see if the speed went up.

So back to answer E...
From how I interpreted it, this new information could actually be consistent with (and hence strengthen) the conclusion .

Lets say we have :
Study 1: 56% (purported cause results in a 56% chance of effect)
Study 2: ruled out 4 improper method
Study 3: ruled out for improper method
Study 4: 34%

So even if we have these 2 points (form study 1 and study 4), it could still support the conclusion that effect disappears with age. Therefore, E doesn't weaken.

Would appreciate any feed back on the reasoning