#16 - Tissue biopsies taken on patients who have undergone

[Administrator]

Complete Question Explanation

Strengthen, CE. The correct answer choice is (E)

Among patients who have undergone throat surgery, frequent snorers were much more likely to
have significant abnormalities in their throat muscles than those who snored rarely, or not at all. The
author concludes based on this premise that snoring can damage one’s throat:

• Premise: Based on biopsies of patients who have undergone throat surgery, those who
  snore frequently are much more likely to have abnormalities in their throat
  muscles than those who snore rarely or not at all.
  
  Conclusion: Snoring can cause damage to the throat of the snorer.

The problem with this conclusion, much like the problem with many flawed causal arguments, is
that the evidence provided speaks only to a correlation. That is, if we were to look at the population
with abnormalities in their throat muscles, and we were to look at the population that snores heavily,
we would see some significant overlap—there would be many people with both throat muscle
abnormalities and heavy snoring habits. This does not necessarily mean, of course, that a causal
relationship actually exists.

Although this conclusion is flawed, in this case the stimulus is followed by a Strengthen question.
There are several ways that we could strengthen such a causal claim (as covered in the PowerScore
LSAT Logical Reasoning Bible):

1. Eliminate any alternate causes for the stated effect
   
   Because the author believes there is only one cause (the stated cause in the argument),
   eliminating one or more of the other possible causes strengthens the conclusion.
2. Show that when the cause occurs, the effect occurs
   
   Because the author believes that the cause always produces the effect, any scenario where the
   cause occurs and the effect follows lends credibility to the conclusion. This type of answer
   can appear in the form of an example.
3. Show that when the cause does not occur, the effect does not occur
   
   Using the reasoning in the previous point, any scenario where the cause does not occur and the
   effect does not occur supports the conclusion. This type of answer also can appear in the form
   of an example.
4. Eliminate the possibility that the stated relationship is reversed
   
   Because the author believes that the cause and effect relationship is correctly stated,
   eliminating the possibility that the relationship is backwards (the claimed effect is actually
   the cause of the claimed cause) strengthens the conclusion.
5. Show that the data used to make the causal statement are accurate, or eliminate possible
   problems with the data
   
   If the data used to make a causal statement are in error, then the validity of the causal claim
   is in question. Any information that eliminates error or reduces the possibility of error will
   support the argument.

Answer choice (A): This answer choice weakens the credibility of the study overall, since selfreporting
might not be as accurate as monitoring (many people might not even realize that they
snore). Since this answer choice calls into question the existence of even a correlation between throat
abnormalities and snoring, it certainly does not strengthen the author’s causal argument that snoring
can damage one’s throat.

Answer choice (B): This choice is entirely irrelevant to the question of whether or not heavy
snoring can have a damaging effect on the muscles of the throat. Regardless of the reasons behind
the throat surgeries, the author’s points to the correlation between heavy snoring and throat muscle
irregularities, and asserts a causal relationship. This choice does not strengthen that causal argument,
so it cannot be the correct answer choice to this question.

Answer choice (C): Regardless of the similarities between test subjects, the point of this stimulus
is to assess the relationship between two factors—snoring and throat muscle irregularities. The
author argues that this reflects a particular causal relationship, and this choice fails to strengthen that
conclusion.

Answer choice (D): This would be a tempting answer choice if the stimulus concerned the
relationship between snoring and throat surgery, but this author is concerned only with the
relationship between snoring and throat muscle irregularities. All of the subjects in this case had
undergone throat surgery, so this choice does not strengthen the author’s argument that snoring can
damage the throat.

Answer choice (E): This is the correct answer choice, because it rules out the reverse explanation
of the correlation between snoring and throat abnormalities (thus meeting the strengthening method
described in point D described on the previous page). If we can confidently rule out the possibility
that the throat abnormalities cause heavy snoring, then this strengthens the argument for the author’s
conclusion that it’s actually the other way around—that snoring leads to the throat abnormalities
discovered in the biopsies.

[netherlands]

Hi there PS,

I went through this question and chose E which happened to be the correct answer choice - but had trouble understanding why D and C couldn't have been strengthening answer choices as well.

In this stimulus those with serious throat abnormalities seemed to be more likely to snore - so the author concludes that snoring causes snoring abnormalities. So, I guess right off the bat this is a causal argument (which I don't think that I really said to myself while going through this question.)

C could tell us that most people looked at probably didn't have any other different health factors that affected them or the throat issue- they were equally healthy with their only differences being their rate of snoring.

D could tell us that the group was fairly representative in that it wasn't skewed from the beginning by being more likely to have hardcore snorers.

E shows that the causal relationship is not reversed. I chose this one because it seemed strongest and to have the greater direct impact on the stimulus.

Can you help me understand why I should have eliminated C and D?

[Jason Schultz]

Hi there netherlands,

The stimulus writer has a serious problem in the data set he/she is using to derive their conclusion: It is drawn only from people who have had throat surgery. Since throat surgery is obviously uncommon and extreme, there is the real risk of bias in the sample. Answer choices C & D actually exacerbate the sample problem and so weaken the author's position.

Answer choice C makes the group even more homogenous and less representative. Imagine if every one was a 65 year old, overweight, lifelong smoker. That makes it impossible to isolate snoring as the cause because everyone is the same.

Answer choice D is also problematic because the group of people who have not had throat surgery is much, much larger than those who have. If people getting surgery are snoring at the same rate as those who do not, how can you connect the surgery to snoring? You would expect the people in the surgery group to be snoring at a much higher rate if there was a causal connection.

Answer choice E is correct because it eliminates the reversed cause-and-effect possibility, which strengthens the author's conclusion. Though it doesn't address the sample problem, it doesn't need to. This is a strengthen question and any help is sufficient to make the answer correct.

[netherlands]

... I must be interpreting the stimulus differently than you are/incorrectly, I'm still not seeing why C and D can't strengthen this argument.

My interpretation is that there is a group of patients who have all had throat surgery - that specific group of patients is examined and its found that within that group of patients ... some snore more than others and thus have different degrees of throat abnormality. As a result - the author concludes that snoring has some affect ( in this case, a damaging one) on the throat - I guess an overall assumption about those in the general population who have a snore and/or have throat abnormalities.

Are you're saying that their receiving throat surgery in the first place sets them apart from individuals that don't receive surgery... If so, I guess I'm not understanding why.

If 25 people have a throat abnormality and don't decide to receive surgery - and 25 people have throat abnormality and do decide to receive surgery... why can't we use the group that does decide to receive surgery in order to make a general statement about all 50 of the people?

Unless we're assuming that the decision to have surgery/or not in the first place is a reflection of a difference in health in which case yes, then it would be biased. But how would we know that? What if some people just don't feel the need to have abnormalities corrected while others do?

I feel like I've again taken another easy question and turned it into something difficult. But grateful for any further explanation!

[BethRibet]

Hi netherlands,

So yes, I think you are probably are imposing a bit more complexity on this stimulus then you need to.

What we know from the stimulus is that among the limited sample of people included in the study -- i.e. those who have had throat surgery, there's a correlation between snoring, and abnormalities in the throat. As Jason noted, the study is already questionable because it is drawing a general conclusion about the throats of people in general based on those who have needed surgery on their throats (sort of like drawing conclusions about people's health choices generally, based only on a study of the choices made by cancer patients -- having an extreme condition or situation makes it unlikely that you can represent anything "in general").

And in addition, it is flawed for a whole other reason, namely that there is only evidence of a correlation between snoring, and abnormalities. We have no information that snoring causes abnormalities, as opposed to abnormalities causing snoring, or something else causing both. E partially corrects this latter flaw, because it rules out one of those possibilities, i.e. that the abnormalities actually caused the snoring. So that's enough to strengthen, though as Jason, it does not make the argument flawless.

C & D do not really address the error made in assuming causation from correlation, nor do they help fix the problem that the sample chosen for the study is not very representative.

The broader questions you're asking are important, and among the sort of things students work through, when taking a course is social scientific research methods. But for the purposes of the LSAT, it will still get you to the right answer if you understand that the LSAT presumes that a very specific or limited sample is flawed, because it does not represent the whole well, and therefore should not be generalized. If you want to explore *why* that's so, I think it's a helpful to study (especially if you're interested in doing research), but beyond the scope of what you really have to do to succeed on the LSAT.

Hope this helps!
Beth

[srcline@noctrl.edu]

Hello

So after reviewing my test, I diagrammed this question as:

Conclusion: Cause: Snoring Effect: damage to the throat.

Premises: Correlation: frequent snorers more likely to have serious abnormalities.

So I had initially picked B. but this is incorrect b/c the patients' throat surgery is irrelevant to the conclusion on how snoring damages the throat , correct?

So for E then, this would strengthen the conclusion because we are trying to strengthen the conclusion that snoring causes damage to the throat and not another reason like abnormalities?

Thankyou
Sarah

[Nikki Siclunov]

Sarah,

This is absolutely correct. One minor correction: don't diagram correlations using an arrow ( ). That's the whole point - you don't know which way the causation goes (if there is any). So, you may be biasing your analysis of the correlation if you're using an arrow to represent it. Remember: a correlation is merely a statistical measure that indicates the extent to which two or more independent variables fluctuate together

Good job analyzing this problem!

[jrc3813]

I have a question about how an answer like D could work if the stimulus was a bit different. Sometimes answer choices like that are the right answer and I want to see if I understand why. In this question both "groups", the snorers and the non snorers, have undergone throat surgery so any correlation between snoring and throat surgery is irrelevant. If snorers are over represented in this study compared to the general population that doesn't affect that fact that within this study snoring and throat abnormalities are correlated. It's like an experiment that makes sure all other variables other than the ones being tested are either removed or constant across both groups.

But if the study in the question was done by dividing snorers and non-snorers into separate groups from the general population then an answer like D could that strengthen the argument. In other words if throat surgery was over represented in the snorer group compared to the non snorer group, there could be an issue of throat surgery being a third cause of both or some other problem.

Is that right? Would a general rule be that if some factor is present across both groups it is irrelevant, but if it's present in one group more than it is a potential problem for the study?

[Eric Ockert]

Hi jrc!

I think the stimulus you describe might go something like:
A random sampling of a certain population found that, among those who snored, a disproportionate number had also had throat surgery. Therefore, snoring must cause some kind of damage to the throat that requires surgery.

If that were the case, then yes, answer choice (D) would be helpful. This fact pattern above shows a correlation between snoring and throat surgery. Although the author concludes that the snoring must cause people to have surgery, it is just as likely that the surgery is causing people to snore. Answer choice (D) would help reduce the likelihood of this reverse explanation, and therefore strengthen the original explanation.

Hope that helps!

[bukkaabh]

I understand the reason why E is the correct answer. However, I was a little thrown off when taking the test as practice because of the first part of the question: "...patients who have undergone throat surgery". Later, in the conclusion, it says that snoring can "damage the throat" of the snorer. I was unsure if the "damage" referred to was as a result of the surgery or the abnormalities. From the explanation, it looks like the damage refers only to the abnormalities, so is it safe to say that the fact that these subjects (patients) have undergone throat surgery is largely irrelevant when approaching the question?