Q21 - Driver: My friends say

[dan]

21. (A)
Question Type: Identify the Flaw
This is a classic "cause vs. correlation" issue. Ownership of a minivan is correlated with lower accident rates, but in no way does this mean that owning a minivan causes one have a lower risk of accident (maybe those who own minivans are just naturally more cautious people in the first place). Thus, the driver is incorrect in concluding so. Answer (A) is correct. An equivalently flawed argument would be:

Many rich people eat caviar. Therefore, if I begin eating caviar I will get rich.

This is, of course, a ridiculous argument because it confuses a correlation with a cause. Eating caviar does not cause one to be rich.

(B) tries to tempt us with the common "sample size" flaw language, but we know nothing about the sample size used in the driver’s research.
(C) is incorrect. This isn’t a case of likely versus certain.
(D) tempts us with familiar conditional logic language: sufficient and necessary conditions. Here’s an example of mistaking a sufficient condition for a necessary condition: "Switching from a sports car to a minivan (sufficient condition) causes the driver to drive more cautiously (necessary condition). Because John has begun driving more cautiously, he must have switched from a sports car to a minivan." No! Not necessarily true. The conclusion in this argument has mistaken a sufficient condition for a necessary condition. This is another way of saying that the argument has reversed the logic. #21 does not contain this type of flaw.
(E) is ambiguous at best. Which source are we talking about? Friends? Research? Either way, we have no way of knowing if they are well-informed or not.


#officialexplanation

[clarafok]

why can't we say that 'low accident rates' = less likely, and lower risk of having an accident if sports car is traded in for a minivan = certain?

thanks in advance!

[bbirdwell]

I'm not sure what you mean. Could you please clarify?

[clarafok]

sorry i totally wasn't being clear in my last post, thanks for following up!

actually, i had chosen C for this question. i now see why the answer is A but i'd like to know why my thinking was wrong. i thought that the driver misinterpreted the fact that since minivans and larger sedans are 'likely' to have low accident rates compared to sports car, it would be 'certain' that by trading in his sports car for a mini van would lower his risk of having an accident.

thanks again!

[bbirdwell]

You're mixing up the facts and getting tangled up in probability language. It is not "likely" that minivans have lower accident rates, it is certain.

The author would correct to say that by driving a minivan, the kind of automobile he drives is less likely to be involved in an accident than the sports car he used to drive. But this doesn't necessarily apply to him. After all, he's a reckless driver.

There is a correlation here between minivans and accidents. It's a logical mistake to conclude that the minivans are causing people to be safe -- this is a classic cause/correlation error. Why? Because there are other factors involved -- it's quite likely that the kind of people who drive minivans (parents) drive more safely. In that case, buying a minivan won't make you a safer driver, as our author concludes. But having kids might.

Do you see what I mean?

[irene122]

I am confused by A because I did not see causlity in the conclusion sentence "so trading my sports car...would lower..." I tought it was correlation since there is no word signs of causality (due to/ because...etc.)

And I have long been confused: does "A lower/ increase B." by itself mean causality?

premise: it has been well established that Vitamin E has positive correlation to heart health.
conclusion: In order to maintain heart health, one should increase vitamin E consumption.

Does the conclusion above imply causality besides conditional reasoning?

[ohthatpatrick]

Good question, Irene.

You're getting to heart of "implied causality", which doesn't actually come with the obvious causal language of "causes, will lead to, because of, due to". Instead, implied causal conclusions already assume that one thing has an effect on another.

The example you gave about heart disease DOES have a conclusion that implies causality.

The phrasing of "in order to get X, you should do Y" implies that doing Y has some effect on whether you get X.

Your conclusion implies that more vitamin E has some effect (some causal influence) on maintaining heart health.

And the point LSAT wants us to see is that from a mere correlation between heart health and vitamin E, we can't assume one has any influence over the other.

In your example, the wording of "lower/increase" has nothing to do with the implied causality. It is the wording of "in order to get X, you should do Y". That wording implies that doing Y has some effect, some causal influence on getting X.

The answer to your question of whether "A lowers B" by itself implies causality is Yes.

Just as in Q21, the verb "lower" IS the wording that implies causality.

If doing X lowers Y, than X has some effect on Y. If something produces an effect, then we call it a cause.

To say that "driving a minivan" lowers the "risk of having an accident" is to say that "driving a minivan" has some effect, some causal influence on one's "risk of having an accident".

If I said that "chewing gum increases your score on geometry tests", then I have again assigned causality.

Let me know if this clarifies things for you.

[saurabhgis]

if A is causing B then does that mean A is a sufficient condition and B is necessary condition?

If the answer to the above question is yes, then does that mean you can apply rules of formal logic like contrapositives, mistaken negation etc. to the cause and effect relationship.

I feel a lot of question stimuli have correlation/causation problem and one of the answer choices address that problem by saying sufficient and necessary conditions are switched. If the conditional logic doesn't apply to correlation/causation issue then that answer choice should be automatically wrong. Right?

I have a tough time wrapping my head around it. Thanks to anyone who can enlighten me a little bit.

21. (A)
Question Type: Identify the Flaw
This is a classic "cause vs. correlation" issue. Ownership of a minivan is correlated with lower accident rates, but in no way does this mean that owning a minivan causes one have a lower risk of accident (maybe those who own minivans are just naturally more cautious people in the first place). Thus, the driver is incorrect in concluding so. Answer (A) is correct. An equivalently flawed argument would be:

Many rich people eat caviar. Therefore, if I begin eating caviar I will get rich.

This is, of course, a ridiculous argument because it confuses a correlation with a cause. Eating caviar does not cause one to be rich.

(B) tries to tempt us with the common "sample size" flaw language, but we know nothing about the sample size used in the driver’s research.
(C) is incorrect. This isn’t a case of likely versus certain.
(D) tempts us with familiar conditional logic language: sufficient and necessary conditions. Here’s an example of mistaking a sufficient condition for a necessary condition: "Switching from a sports car to a minivan (sufficient condition) causes the driver to drive more cautiously (necessary condition). Because John has begun driving more cautiously, he must have switched from a sports car to a minivan." No! Not necessarily true. The conclusion in this argument has mistaken a sufficient condition for a necessary condition. This is another way of saying that the argument has reversed the logic. #21 does not contain this type of flaw.
(E) is ambiguous at best. Which source are we talking about? Friends? Research? Either way, we have no way of knowing if they are well-informed or not.

[hyewonkim89]

My question is similar to saurabhgis'.

I chose D as an answer and I still have a very hard time understanding your explanation on D in terms of sufficient and necessary conditions.

How is "Switching from a sports car to a minivan" a sufficient condition and how is "causes the driver to drive more cautiously" a necessary condition?

Thanks!

[ohthatpatrick]

Good questions.

Causality CAN be represented conditionally, and, yes the cause would be the sufficient and the effect would be the necessary. And, yes, you could judge that with the same sort of contrapositive, mistaken negation, etc. type thinking we use for all conditionals.

Here is a conditional logic flaw that is dressed up in causality:
Receiving flowers puts Betty into a good mood. Since I just saw Betty in a good mood, she must have received flowers recently.

In conditional form we had:
prem 1: Flowers --> Good Mood
Prem 2: Good Mood.
Conc: Flowers.

This is a necessary/sufficient flaw. The author's logic is reading that original conditional statement backwards. He's thinking "since good mood --> I conclude flowers". We know that's a mistaken reversal.

Whenever you see a flaw answer choice describing the necessary/sufficient flaw, just ask yourself this:
"Did the premises contain a conditional statement?"

If not, get rid of that answer.

If so, see if the author interpreted that conditional statement illegally in his move from premise to conclusion.

In Q21, there is no conditional language in the premise, so choice (D) is bogus.

The previous poster was just trying to show you what a Sufficient/Necessary error COULD have looked like, using the same topic of sports cars / minivans / accidents etc.

His whole point, ultimately, though was that there was nothing like that actually going on in this argument.

Final point: make sure you realize that a correlation to causality flaw has nothing to do with Nec/Suff conditional logic flaw.

In correlation to causality we're essentially getting
prem 1: A happened.
prem 2: B also happened.
Conc: A caused B

The flaw here is that just because A and B happened to coincide, that doesn't mean that there is a causal connection between them.

If your evidence contains a correlation, you're almost NEVER going to see a conditional logic flaw be the problem.

By contrast, a Nec/Suff conditional logic flaw looks something like:
prem 1: A --> B
prem 2: ~A
conc: ~B

Again, just remember that you need a conditional rule as a premise in order to have a Nec/Suff flaw.

Let me know if there are still any questions.

[dli]

I've been reading this question over and over again. While I understand why A is correct, somehow I am still having trouble with C.

I understood the question as:
it is certain that minivans and larger sedans have very low accident rates compared to sports cars. trading one's sports car in for a minivan would lower one's risk of having an accident. but since the driver drives recklessly it probably won't affect him.

i thought that maybe the type of car one drives is only one of the few factors that affect one's risk of having an accident (another factor could be how carefully one drives). so it is not certain that switching cars would definitely lower one's risk but the driver thinks because it is likely he is certain to have less risk having a minivan or a large sedan. so isn't this pointing at answer C? I somehow see A in my explanation above too... -_-
Thank you!!

[richn12]

I thought that there was a conditional logic in the premise but it was also presented as a cause and effect as well. The reason that I thought (D) was wrong was because the conditional relationship in the conclusion was not a mistaken reversal, it simply restated the original condition. If the conclusion was "by having a lower risk of accident, i traded my sports car for a minivan" wouldn't this make (d) correct??

[ohthatpatrick]

The only thing in the premise that could be construed as conditional logic would be the first sentence.

Since the friends say that the driver WILL one day have an accident, we could represent that as

Keep driving sports car recklessly --> one day have accident

But the conclusion doesn't use either of those ideas, so there's no way that the conclusion is guilty of Nec/Suff (no matter how we re-write the conclusion).

"Lowering the risk" of an accident is not a negation of "WILL one day have an accident" (the negation would be "will NOT one day have an accident").

And trading in sports car for minivan is not a negation of "do NOT drive sports car recklessly". I'll admit, it's closer. If you no longer have a sports car, it's hard to drive it recklessly.

But it's hard to interpret from the first sentence what the friends were saying WILL lead to an accident:
- the sports car
- the reckless driving
- or the combination?

The 2nd sentence is not a conditional statement, unless you're trying to turn every sentence into a conditional statement by using the verb.

We could say the 2nd sentence is:
Minivans or Larger Sedan --> Lower accident rate than sports car

But in order to commit a Nec/Suff flaw, you would have to argue that "if you have a lower accident rate than sports car, then you are a minivan or larger sedan".

I see how you might be thinking that the last sentence, if reversed would sound like
"lower accident rate than sports car --> minivan".

But that's confusing the adjective "lower" with the verb "lower".

The 2nd sentence is using the adjective "lower". The 3rd sentence is using the verb "lower".

It is a fact that Death Valley, where cactus grows, is at a LOWER altitude than Sonoma Valley, where cactus does not grow.

But that is not the same as saying that "growing cactus LOWERS your altitude".

The key difference in meaning between the 2nd and 3rd sentence is both why (A) is the answer and why (D) is not. Because we're shifting from a simple numerical fact to a causal claim, we can't be accused, as (D) is saying, of taking the same two ideas but putting them in reverse order.

We are CHANGING one of the ideas from descriptive to causal.

In regards to the previous poster asking about (C), there is nothing about "a RESULT being likely" in the premise. The friends are saying in the evidence that the result of having an accident is CERTAIN (it will one day happen).

The statistic about minivans/sedans/sports cars has no causality to it, so there is no sense of a RESULT being implied.

If I say that "prisoners in US prisons have very low rates of speaking French compared to prisoners in French prisons." Does that mean that "going to a US prison makes you less likely to speak French than not going to a US prison"?

Is that evidence that if you go to a US prison, it is LIKELY that you will not speak French as a result (but there may be other factors that influence it as well)?

No! Going to a US prison probably has no effect on whether or not you do or will speak French.

The conclusion DOES speak about a result being certain. It thinks that changing to a minivan is CERTAIN to lower the risk of accident.

Can we say that the premise merely said that "changing to a minivan is LIKELY to lower the risk of accident"?

No. It's totally reasonable for you to think that might be a true statement in the real world, but that is not a textual match for our 2nd sentence in this argument.

Hope this helps.

[WaltGrace1983]

The way that I understand this one is simply that (C) exhibits a different flaw that is simply not here. A (C) type flaw would look like this: "It is likely that working hard will lead to increased pay. I have been working really hard lately. Thus, I am getting a raise soon!" The conclusion shows that something - getting a raise - is absolutely 100% bound to happen merely from the premise that I have been working hard lately. Now working hard is LIKELY to lead to a pay raise; but it doesn't HAVE TO. I am assuming that it does.

That is how I eliminated (C) and I do this a lot in flaw questions actually. There are only so many logical flaws and, typically, the LSAT likes to just recycle flaws that are not applicable to the question but have actually appeared in other questions. For example, (D) shows up all the time in other LSAT questions, but not this one.

Also, who's to say that was is true of the group is true of this one specific person? Even if no has ever gotten into an accident when driving a minivan, that does NOT mean that the minivan is the cause. When we assume that driving a minivan will lower the risk of an accident, we are assuming that the minivan in some way causes something. It may not.

[contropositive]

From reading the posts above, it looks like C is wrong because the argument doesn't say "minivans are likely to have lower accident rates" am I right? if that type of flaw came up, it would have the word "likely" in the premise?

[ohthatpatrick]

Yes, in order for (C) to be correct, the premise needs some mention of a likely result in the premise. It doesn’t have to be the word ‘likely’ precisely, but some equivalent like
most of the time
tends to
usually
probably
generally

The structure of a choice like (C), that says the author misinterpreted evidence that X as evidence that Y would basically force us to match X up with premise and to match Y up with the conclusion.

Does the conclusion say that a result is certain?
Yes.

Trading sports car for minivan WOULD lower accident risk.

So is there a premise that says “trading sports car for minivan WOULD PROBABLY lower accident risk?”

No. There’s statistical correlation that minivans have lower accident rates, but if we interpret that statistic as “if I were to switch to minivan, I’d probably have a lower accident risk”, then WE are committing the causal flaw.

If we say “rich people are more likely to eat caviar than are poor people”, are we saying that “if we started eating caviar, we would LIKELY get richer”?

Hope this helps.

[hyk1310]

Dear Patrick,

In explaining implied causality, you mentioned the following example "chewing gum increases your score on geometry tests."

If causal statements can be represented with conditional statements, then would chewing gum --> increasing score make sense?

If you chew gum, then it increase your score.

But if you switch out the if, then to in order to, must,

the following statement occurs:
In order to chew gum, you must increase your test score. This doesn't seem very intuitive to me. Could you explain this? Thank you in advance

[ohthatpatrick]

Sure, so don't switch it.

No one is really suggesting that we think about this conditionally in the first place. Students often want more of the test to have the mathematical certainty of conditional logic, but it's not a good tool for most of the thinking we should be doing.

You're getting confused because a given conditional statement,
A --> B
can be representing tons of different ideas

A causes B
if A, then B
All A's are B's
A requires B
in order to have A, you must have B
Until you have B, you won't have A
etc.

So the fact that we turned "chewing gum increases your score on quiz" into
"Chew gum --> increase score" doesn't mean that we'll suddenly like the same conditional rule applied to any other language formulation:

chewing gum requires that you increase your score
all gum chewers have increased scores
until you've increased your score, you haven't chewed gum

These are crazy!

That's not how conditional logic is used. We can go from lots of different phrasings into one consistent conditional logic symbolization, but we can't reverse that and expect it to make sense.

All lions are carnivores = Lion --> Carnivore

but I wouldn't expect it so make intuitive sense to say stuff like:
Being a lion causes you to be a carnivore
In order to be a lion, you must be a carnivore
Until you're a carnivore, you're not a lion
etc.