QUESTION TEXT: Tissue biopsies taken on patients…
QUESTION TYPE: Strengthen
CONCLUSION: Snoring can damage your throat.
REASONING: There’s a correlation between snoring and throat damage.
ANALYSIS: Correlations can’t prove anything. Whenever there’s a correlation, there are four possibilities:
- Snoring causes throat damage.
- Throat damage causes snoring.
- A third factor causes both.
- It’s a coincidence.
If an argument uses a correlation to prove causation, you can strengthen the argument by eliminating one of the other possibilities.
Throat surgery is mentioned in a couple of answers. But it’s irrelevant. The conclusion is about snoring in the general population, not just those who had surgery. Surgery is only mentioned because it’s what let us discover abnormalities.
___________
- This weakens the argument by showing that the study’s data may be flawed.
- It doesn’t matter why patients had surgery. Surgery is just how we established there was throat damage.
- This is a common wrong answer. Studies don’t need to use subjects of the same age/weight etc. The study is about snoring in the general population, so the argument is stronger if the study used subjects from the general population.
- Throat surgery doesn’t matter. Throat surgery is the reason we discovered abnormalities, but the argument is about the abnormalities in the whole population, even those who don’t have surgery.
- CORRECT. This answer eliminates the second possibility from my list above. This strengthens the conclusion that the first possibility is correct.
Recap: The question begins with “Tissue biopsies taken on patients”. It is a Strengthen question. To practice more Strengthen questions, have a look at the LSAT Questions by Type page.
Free Logical Reasoning lesson
Get a free sample of the Logical Reasoning Mastery Seminar. Learn tips for solving LR questions
MemberHaile P. Selassie says
(E) does touch on one of the four possibilities of correlation that you mentioned (reverse-causation), so I know why (E) is right but I had trouble pinpointing why (C) is wrong.
You wrote for (C) “Studies don’t need to use subjects of the same age/weight etc. The study is about snoring in the general population, so the argument is stronger if the study used subjects from the general population.”
This means that the more heterogeneous a sample is, the better. However, I think that there’s a critical difference between a study that aims to describe certain features of a population and a study that aims to prove causation based on correlation.
In a study that describes a population, you are right that we should use subjects that match the composition of the general population, i.e. heterogeneous sample. An example is a study that proves whether Americans prefer apple pies or pumpkin pies. In that case the more heterogeneous the sample is, the more faithfully it would reflect Americans’ (i.e. the population’s) actual preferences, hence the better.
However, a different approach should be taken when proving causation based on correlation data (correct me if I’m wrong). In this case homogeneous sample is better because It is important to control for such variables as age and weight, so studies do need to use subjects who are of similar age/weight etc. (but only to the extent that these variables are causally linked to the dependent variable) Hence an answer to a classic Weaken question whose stimulus says “snoring is correlated with throat damage, hence snoring causes throat damage” would be something like “(A) snoring is correlated with being 65+ years old, and old age causes throat damage.” Here, old age would be the “third variable” or “confounding variable.” Problem here was that the sample was too heterogeneous.
To strengthen this argument, we would need to create two or more separate samples based on age: those under 65 and those over 65 (note that each of these two samples would thus be homogeneous). Then the argument is strengthened if correlation is observed within each group because now we know that age differential has nothing to do with the correlation.
Now, let’s return to answer choice (C) of Q16: “All of the test subjects were of similar age and weight and in similar states of health.” Unlike the hypothetical Weaken question I solved above where I explicitly said “old age is known to cause throat damage”, here it is unclear how age or weight or “states of health” is related to throat damage. Maybe teens are more vulnerable to throat damage or older adults are more vulnerable. Maybe underweight people are more vulnerable or overweight are more vulnerable. And “state of health” is just so broad and vague. Cardiovascular health? muscular health? throat health? etc. Maybe healthier people are counter-intuitively more prone to throat damage, etc. At any rate, we need to add these additional assumptions that the answer choice itself does not hint at all (correct me if I am wrong).
My guess is that generally speaking, homogeneous sample like (C) does strengthen a correlation-causation argument, but there must be something else that’s wrong with (C). Maybe you can clarify further.
FounderGraeme Blake says
Interesting idea. It’s true that there is always a possibility that a third factor causes both snoring and abnormalities. As I understand it, you’re arguing that C reduces the odds of such a third factor.
There may be something to that. I guess in that case I would fall back on the idea that you must pick the stronger answer, and E is definitely stronger. C merely reduces the odds of a possibility. Whereas E directly eliminates something that would destroy the argument.
I’d have to think more about your questions to know whether it’s correct, but you may be on to something. However, I do not think it is as strong as E.
S says
Another question – what if this had been a weakening question? Then would the correct answer choice be one that INCLUDES one of the four possibilities instead of eliminating them?
FounderGraeme Blake says
Yes, the correct answer would likely have pointed out possibility #2 or #3. Though of course they could weaken any way they want.
Note: This is an old comment but I wanted to clarify the point.
S says
What if E had been: Snoring does not cause the abnormalities in the throat muscles discovered in the study. Then that would have been a direct contradiction to the conclusion, meaning it would be incorrect, right? It’s why I eliminated E. I didn’t read the answer choice carefully and I chose D instead.
FounderGraeme Blake says
Yes, that’s a reversal and would have been incorrect. Always vital to see which way statements go.
Note: This is an old comment but I wanted to clarify the point.