Q11 - Recent studies indicate a correlation

[andrewgong01]

Question Type:
Flaw


(A) Yes! If some but not all types of damage to ch6 lead to schizophrenia, does that weaken the author's argument? Of course. The fact that some types of damage DO lead to schizo basically refutes the conclusion. And the fact that some types of damage DO not lead to schizo explains the exceptions in the author's evidence in which people do have damage but don't get schizo.




(C) Tempting a bit, since the author seems to reach a hastily broad conclusion. But there's no way to say the sample is "unrepresentative". It seems like she is considering cases of people with/without damage to ch6 and with/without adult schizophrenia. Her bad move is not that she assumes these people are representative of all cases; her bad move is thinking that since damage to ch6 and schizo are not invariably associated, they must have no causal connection.



#officialexplanation

I switched from "A" to "C" in Blind Review and Now I am more convinced that it is "C".

I remember reading a different post from a different instructor who said that when you see a study being cited you want to check for potential population sample bias. In this case, how did we know though that the sample population was not biased because all it says is "some people" the author "knows" and for all we know this "some" person could be like two people. I understand that the mistake the author makes is then going on saying there is no causation because there are cases where the correlation does not hold but to reach this step in reasoning doesn't the author do what "C" describes where it depends and draws on the exception cases? In other words, the author goes from saying there isn't like some 100% correlation which means that there is no causation but this premise is supported from the "some people" the author knows?

The other reason why I switched from A to C is that I noticed on review the stimulus also said there are people the author knows that has schizophrenia but do not have the chromosome damage. However, this part i understand now from the post above in that the existing research on correlation never said the damage MUST all the time cause it and, more importantly, be the only cause of schizophrenia.

[KenM242]

Well, the answer is already (A), so I know whatever I thought and think is wrong BUT I need someone to help me get over (E), PLEASE.

What does (E) mean exactly?

This is how I interpret it: The author thinks that because there is a correlation between schizo and chromosome dmg, there has to be a causal relationship between the two.

Am I not understanding what 'imply' means here?

I am sorry if this question ends up turning into an ESL class topic.

[WesleyC316]

I initially chose E because I thought the author was trying to deny the correlation stated in the first sentence but ended up denying a causal relationship, but later I realized that's not necessarily what the argument was doing.

Also, plenty of people seem to be bothered by the wording in the argument and got tempted by C. To me C is the most obvious wrong answer choice. The author is trying to deny a causal relationship, so all he has to do is to provide a counter example, just as he does in the argument. One incident is enough to negate a causal relationship, so it's really not the matter of unrepresentative sample here.

#Edit# Now that I think about it, one counter example doesn't work. It could be some other factors negating the impact of that cause. But C is still incorrect.

[snoopy]

I chose E and still can't see why it's wrong. When I first read the stimulus, I read that because there are SOME people with damage to ch6 but no schz and SOME people w/o damage to ch6 but have schz, the implication is there is no correlation and, hence, no causation. Did I make the mistake of assuming that the author implied no correlation between chromosome 6 and schizophrenia?

I can't ever see correlation equaling causation, so the explanation (for A) saying that there could be a causal relationship can't stick with me. In statistics, correlation never equals causation. Hence, I chose E because it seemed like the author implied there was no correlation and, hence, no causation.

But, in my head, I justify A like this: because there are SOME damages in ch6 that lead to schz, there is a correlation. Just a weak one.

But, then, I would argue that, well, even if there is a weak correlation, it does not imply a causal connection.

What is wrong with my thinking here?

[ohthatpatrick]

There are a few things:

1. you can't read (E) as "the author assumes [a lack of] correlation implies [a lack of] causation."

And we can't read it as "the degree of correlation implies the degree of causation".

We can only read it as, "the author thinks that a correlation between two things implies a causal relationship between those two things".

More importantly, there IS a correlation cited in the first sentence. So when (E) refers to a correlation, it's referring to that explicitly cited correlation.

And the author certainly doesn't assume, from the fact that damage to 6 and schz are correlated, that damage to 6 causes schz.

------------------

2. The existence of some counterexamples doesn't invalidate the idea of a correlation between two variables.

I haven't taken statistics, but my understanding is that correlations can be expressed in terms of how strong the relationship is, with the idea that in all but the most extreme case (a perfect correlation), there are always data points that do not fit the trend.

So when the author brings up the existence of counterexamples, he isn't saying "there is zero correlation". If he were to conclude that, then we could accuse him of the flaw of thinking that "in order for there to be a correlation between two variables, there can't be any counterexamples".

What he is doing is assuming that "In order for there to be a causal connection between two things, there must be a perfect correlation between those two things".

---------------

3. Choice (A) isn't talking about correlations. It's talking about causality. The verb "lead to" indicates a causal connection.

So (A) is literally saying "the argument ignores the possibility that [only some types of damage to 6 have a causal connection to adult schz]

[A B]

Hi, so I have a question regarding choice B. I see what everyone in this thread has said about the argument's conclusion states that there's no causal connection, so this is the opposite of that, but I want to understand why my reasoning was wrong.

I was between A and B and I thought that A didn't ultimately work because it didn't account for the fact that some people who don't have damage to chromosome X at all, nevertheless still have schizophrenia. But I guess I see from Patrick's examples that some instances of cause absent but effect present is okay.

When I read the argument I thought, no, this person is wrong, there could still be a causal connection where chrom6 leads to schizophrenia, just that it only happens when it is in conjunction with chrom7, for example. Something showing that chrom6 leads to schizophrenia but in some way the author didn't account for. A I ruled out ultimately for the reason I said above, but more problematic is why I chose B:

I thought B was doing what I predicted! I thought, oh look, this is the flaw in the argument, this is her mistake! The author presumes that schiz is caused solely by chrom6 (she doesn't account for the fact that it could be caused by chrom6+ something else, together leading to schiz) and so she draws the faulty conclusion that there is no causal connection there because she presumes this faulty assumption that schiz is only caused by chromosomal damage.

Does that make sense to you? Hoping you can help me figure out why that's wrong!

Thanks

[JeremyK686]

Moves Made:
Move 1: assert that two items are functionally associated
Move 2: present information that establishes some disassociation
Move 3: claim that the two items are not associated.

Analysis:
The support isn’t extensive enough to draw such an extensive conclusion. Only particular premises can’t support a universal statement. It’s a violation of rule eight and two in the rules of the syllogism. From particular premises, nothing follows. The conclusion can never be broader than its premises. This conclusion is too strong.

Maybe there’s some causal connection there, despite the facts given.

Answer Choices:
(A) This lets me know that there are certain types, at least one type of chromosome damage that leads to schizophrenia. Is it adult schizophrenia or is it children’s schizophrenia. I feel like it’s just as much right as it could be wrong. It could be saying that damage could lead to schizophrenia (in general, of any type). This allows the evidence to exist while providing evidence against the conclusion.

(B) This is likely opposite of what the argument assumes. If this is true, then something else could’ve caused such schizophrenia. More explicit, this answer is concerned about damage causing schizo while the argument is about damage and schizo having no causal relationship.

(D) This mistake isn’t made in the argument above. This contradicts the argument because the argument asserts that there is no causal relationship at all.

(E) The argument goes from a correlation, to some supporting fact, and then arrives at no causation. It’s more likely that the argument is assuming that correlation doesn’t imply causation.

(C) Knowing the particularities of induction errors is what this question set taught me and what this answer forced me to consider. For me, getting through answer (C) was tough because it expresses a flaw that is within the same general class as the flaw committed by the argument. The argument’s flaw and this flaw are within the ‘induction error’ class.

I feel like the argument commits the fallacy of exclusion whereas the answer choice suggests that the argument makes a sort of hasty generalization error. They have specific differences...

Fallacy of Exclusion: overlooks that the given quality is common/possible amongst different groups/types.
The Move: I studied some and made a generalization about all.
Example: Some people who drive drunk have never gotten into an accident. Therefore, driving drunk and getting into accidents are not causally related.
Objection: Maybe those people who drove drunk only had to drive down the street. There are some drunk drivers with tougher routes who have gotten into accidents.

Hasty Generalization: assuming that the quality is common to all/most members of the particular group studied (unlike the exclusion fallacy, this isn’t concerned with other groups/types).
The Move: I only studied one group and failed to look for the same pattern in other groups.
Example: Some republicans are racist. Therefore, all republicans are racist.
Objection: My mother is a republican and she’s not racist.

I hope this helps and I hope my thoughts are within reason!