Q11 - Studies have shown that those who

[ohthatpatrick]

Question Type:
Match the Flaw

Stimulus Breakdown:
Conclusion: If you contract a cold, you probably don't take daily vitamin C.
Evidence: People who take daily doses of vitamin C are less likely to contract colds than those who don't.

Answer Anticipation:
Because the first sentence is a correlation "people who are A are less likely to be X than are people who are not A", we might assume the flawed conclusion will be causal. But the conclusion is just statistical as well. So what is the statistical flaw?

If we know that "People who are A are less likely to be X than people who aren't A", can we conclude that "If you are X, then you are probably not A?" It would probably be easier to hear the flaw if you switch between two equivalent forms: "Most X's are not A" and "If you are X, then you are probably not A". Have we proven that "MOST people who contract a cold do not take vitamin C daily?" No.

Consider this analogy: "People who are right handed are less likely to like Jimi Hendrix than are people who aren't right handed. Thus, if a person likes Jimi Hendrix, they are probably not right handed."

The first sentence is saying that Jimi (a left handed guitarist) is more popular among lefties than righties (maybe 80% of lefties like him whereas only 65% of righties like him). Does that mean that if I find a Jimi Hendrix fan, they are probably left handed? Does that mean "most Jimi Hendrix fans are left handed"? No. Even though Jimi is more popular among lefties, there are 8 times as many right handed people as left handed people in the world. So most Jimi fans are still right handed.

This is a tough math flaw to figure out, but we can map out the structural parts and just look for an answer to match those:
prem - People who are A are less likely to be X than people who are ~A.
conc - If a person is X, they are probably ~A.

Correct Answer:
E

Answer Choice Analysis:
(A) The first sentence is not a more/less likely correlation, so I would stop reading there.

(B) The first sentence is not a more/less likely correlation, so I would stop reading there.

(C) The first sentence says "rats given low cal were more likely to live longer than rats not on low cal". So to match, we would need a conclusion that says, "If something didn't live as long, it was probably not on a low cal diet". Instead it transfers advice from rats to humans, creating a Comparison Flaw, not a statistical one.

(D) The first sentence is not a more/less likely correlation, so I would stop reading there.

(E) YES. "Cars that have regular oil changes are less likely to have problems than cars who don't have regular oil changes". We would need a conclusion that sounds like, "So if a car has a problem, it's probably a car that doesn't have regular oil changes". And the conclusion is pretty close to saying that!

Takeaway/Pattern: This is a nice reminder that we can be stumped as to the exact nature of the Flaw but still successfully pull out a correct answer on Match the Flaw, if we pay attention to the Type / Strength / Number of claims involved in the original argument. Being able to label the premise a "correlation" allowed us to get rid of 3 mismatching answers. The conclusion has the strength of "probably", so that also would have allowed us to get rid of certain conclusion like (A) and (D).

#officialexplanation