Q25 - A recent survey showed that

[noah]

The conclusion of this flawed argument is that more people think elected officials (to keep it simple, I'll just say "officials" or "politicians") should resign if indicted than believe they should resign if convicted. (I'd expect it to be the other way around!).

What's the support? Some survey results, of course! 50% believe politicians should resign if indicted. What do you expect next? You expect some number, like 35%, believe that politicians should resign if convicted. But, instead, we learn that 35% believe that politicians should resign ONLY if convicted. Strange. What's the flaw then?

Well, these groups aren't directly comparable. The total number of people that believe politicians should resign if convicted might be much larger than 35%. 35% is the number of politicians that believe that politicians should resign ONLY if convicted, but perhaps it's 90% of the population that believe they should resign if convicted (the 55% - 90-35) would believe politicians should resign if convicted and also if something else - like a sex scandal - occurs.

Still confused? With this argument, it's helpful to put it into real numbers:

100 people were surveyed:

50 said they think officials should resign if indicted
35 said they think officials should resign ONLY if convicted

Which groups are missing?

We don't hear how many think officials should resign if convicted (as opposed to only if convicted).

In order to make the comparison the conclusion draws, we would need to know that last figure.
(B) is a complex way of describing this flaw. If we were to write out each premise in conditional logic, we would have:
Indicted --> Resign & Resign --> Convicted.

The key is to see that the second statement is not Convicted --> Resign, in which case the conclusion would be true. By the way, "Convicted --> Resign" would translate to "elected officials should resign if they are convicted of a crime", and we have a different statement because of the only, which makes that part the necessary part of the logical statement.

(A) is incorrect because it refers to sample size, which is irrelevant.
(C) is incorrect because there is no ambiguity.
(D) is tempting because it refers to there being two different queries, however, there is no conclusion about a specific belief! The conclusion is a comparison.
(E) is unsupported -- the premised could all be true; they do not contradict each other.

Does that clear it up?

[peg_city]

I'm sorry, but I'm not at all getting what B is trying to say

What is the difference between a sufficient condition and a required condition? Isn't it the same thing?

What is the sufficient condition in this question?
What is the required condition in this question?

I fell for D, which, after your post, I now understand is wrong.

[noah]

A sufficient condition is one which is enough to ensure that something (the required) occurs. While the required is the thing that must occur if the sufficient condition is met (or occurs).

For example: I will laugh if apple juice comes out your nose.

This can be represented as apple juice nose --> I laugh

The apple juice nose is the sufficient condition (it's sufficient to make me laugh), and me laughing is necessary or "required" (it's necessary that I laugh if apple juice comes out your nose).

Confusing required/necessary and sufficient in this example would be to say "If I laugh, it must be that apple juice came out your nose."

While that might be the reason, there could be other reasons. Get it?

In the LSAT question in discussion in this thread, the conclusion treats resign --> convicted ("you should resign only if convicted") as if it actually was convicted --> resign ("you should resign if convicted" - or, to put the sufficient first, "if you're convicted, you should resign").

This is a pretty key idea, so if you don't have our LR strategy guide, you should probably grab it and read the conditional logic chapter.

I hope that clears up this for you.

[skapur777]

What does it mean when they say

"confused X with Y"? does that mean the speaker thought it was Y but it was really X? if so, isn't answer choice B backwards? They thought it was a sufficient condition but it was really a necessary condition:

so i thought it should read "confuses a necessary condition with a sufficient condition"? They thought it was a sufficient condition, based on the last line, but it was really a necessary condition. thats where the flaw is

[noah]

What does it mean when they say

"confused X with Y"? does that mean the speaker thought it was Y but it was really X? if so, isn't answer choice B backwards? They thought it was a sufficient condition but it was really a necessary condition:

so i thought it should read "confuses a necessary condition with a sufficient condition"? They thought it was a sufficient condition, based on the last line, but it was really a necessary condition. thats where the flaw is

Good question! You've got the flaw right.

The wording is tricky, but I think with that grammatical construction, the LSAT can go either way. "She confused baking soda with salt" would make sense for either mistakenly thinking it's baking soda or salt.

I've found "mistakes X for Y" to be more specific (it's really X, but it was mistakenly considered Y - The Man Who Mistook His Wife for a Hat is how I remember that).

[mcrittell]

In the LSAT question in discussion in this thread, the conclusion treats resign --> convicted ("you should resign only if convicted") as if it actually was convicted --> resign ("you should resign if convicted" - or, to put the sufficient first, "if you're convicted, you should resign").

I have no idea how to diagram the conclusion (given all the "if"s), so it's hard for me to visualize the flaw and, as you say, how what the arg tries to do.

[noah]

Perhaps you'd get something out of playing the LSAT Arcade Game "If/Then"

In general, "only if" means something is necessary, while "if" means it's sufficient. The LSAT can definitely play with those rules, so don't become dogmatic in your application of them.

[mcrittell]

In the LSAT question in discussion in this thread, the conclusion treats resign --> convicted ("you should resign only if convicted") as if it actually was convicted --> resign ("you should resign if convicted" - or, to put the sufficient first, "if you're convicted, you should resign").

I have no idea how to diagram the conclusion (given all the "if"s), so it's hard for me to visualize the flaw and, as you say, how what the arg tries to do.

I understand suff/necessary conditions, but I don't know how to diagram "more ppl believe tht elected officials should resign if indicted than believe that they should resign if convicted"---which's why I' having trouble understanding your explanation.

[noah]

Ah! Got it.

When they say that X% believe politicians should resign if indicted, we can focus on the "if" and that - as is often the case - is the sufficient (big exception is when it's "only if"). So, if indicted, then resign. It's possible those folks believe there could be other reasons to resign (so an indictment isn't necessary), but, all we care about is that an indictment is one of them. So:

indict --> resign

The other statement, about politicians resigning ONLY if convicted, that's the only reason in the world. But, if a pol. is convicted, there still might be a way they shouldn't resign. But, if they resign, the only reason is a conviction. So:

resign --> conviction

Does that clear it up?

[mcrittell]

I guess I don't understand how I-->R/R-->C fit in because I don't understand the conclusion, the stim's last sentence.

[noah]

It's comparing the number of people that would say "Resign!" if a politician were indicted versus if a politician were convicted.

But, the percents given don't allow us to compare those numbers, since the 35% isn't telling us how many would call for a resignation based on a conviction - that's the number of people who must see a conviction to call for a resignation. There might be many more people who think a conviction should lead to resignation, but they are also open to other reasons.

Analogously: 20 people are surveyed about when they will give money to a charity.

15 of them say they will if a friend is involved.
5 say they will give money ONLY when the charity helps their family.

It seems like there's more people who will give money if a friend is involved, but the question that hasn't been asked is "will you give money if a family member is involved?" and the answer to that might be 20 (100%).

[sr]

I thought the answer was A because:

The conclusion is saying that "more people believe", however we don't really know what more people believe. We only know what more people POLLED believe. The entire thing was a poll (a sample of the general population) and the conclusion is making a claim on the general population.

It could be true that in the poll more people of those polled believe one thing, but in the actual population more people believe something else (since the people polled are just a subset of the population).

[mcrittell]

Agreed w the above poster, I thought the flaw was A. Thgts?

[anjelica.grace]

Agreed w the above poster, I thought the flaw was A. Thgts?

I think (A) is wrong because drawing a conclusion about the population in general based only on a sample, on its face, is not a flaw. It's perfectly acceptable to generalize based on a sample so long as it's representative of the whole, and we are not given any reason to doubt its representativeness.

Usually, I find that sampling flaws are the credited response when the stimuli hints at un-representativeness such as a scope shift in groups (ex. poll of voters in certain states taken to represent the political views of the whole country). Otherwise, if there's no indication of such, drawing a sample alone is not a flaw.

[ManhattanPrepLSAT1]

I think (A) is wrong because drawing a conclusion about the population in general based only on a sample, on its face, is not a flaw. It's perfectly acceptable to generalize based on a sample so long as it's representative of the whole, and we are not given any reason to doubt its representativeness.

Perfectly stated! On the LSAT, unless we're given a reason to suspect whether a survey is representative, we accept that it is.

[hwangbo.edu]

I think (A) is wrong because drawing a conclusion about the population in general based only on a sample, on its face, is not a flaw. It's perfectly acceptable to generalize based on a sample so long as it's representative of the whole, and we are not given any reason to doubt its representativeness.

Perfectly stated! On the LSAT, unless we're given a reason to suspect whether a survey is representative, we accept that it is.

Q25 fails to meet LSAT testing standards. Answer A should not have been included as a non-credited response (non-CR) because it does address an actual flaw in the reasoning - and the stimulus clearly indicates a difference between premise and conclusion. Specifically, the premise specifies "people polled" to describe the population surveyed, yet the conclusion generalizes to "people." Certainly many LSAT CRs have turned on the difference of a single word - but in those cases, there was arguably no other correct alternative CR.*

To further the case in point, add one word to the conclusion: "Therefore, more people believe that elected officials should resign if indicted than believe that they should resign only if convicted." Adding this one word (changing 'if' to 'only if') ensures the conclusion matches its premise - thereby fully correcting the flaw identified in Answer B (CR). And suppose this was the case, ask yourself, how could anyone now argue that Answer A is incorrect?

The LSAT standard is that each stand-alone CR is 100% justified by the stimulus and, correspondingly, the four non-CRs are 100% unjustified. Furthermore, the nature of any flaw question is yes-or-no (vs. "best answers the question"). Like Yoda says, flaw answers "do or do not, there is no try" on their own either describe a flaw or not. Q25 has two correct CRs, each which can stand on its own, and therefore should be withheld from scoring.


*SIDE NOTE: Also, in at least modern LSATs, questions involving surveys often give clues that support their validity (as being representative) so as to avoid this very confusion. Would anyone have any other example Qs in which the stimulus does or does not conform to this observation?

[WesleyC316]

This must be one of the most despicable questions of LSAT. The tempting answer choice A, and the time pressure given that it's Q25. I fell for it without checking B to E at all.