#16- When surveyed about which party they would like to see

[mokkyukkyu]

Hi, I was wondering...
So "a group" in C means the respondent right?

Thanks

[Nikki Siclunov]

This is correct. The argument commits a classic Error of Division:

• Premise: According to a survey, 40% of respondents are Conservative, 20% are Moderate, and 40% are Liberal
  Conclusion: Most respondents want to see a legislature that is 40% Conservative, 20% Moderate, and 40% Liberal

Of course, there is no evidence that the preference of each respondent is consistent with the overall survey result: we're assigning a property that belongs to the entire class of things to each individual member of that class. It's entirely possible that a Conservative respondent would want to see a 100% conservative legislature, just like a Liberal respondent would want to see a 100% liberal legislature. In other words, the argument takes for granted that the preferences of a group as a whole (the respondents) are the preferences of most individual members (i.e. each respondent) of that group.

Hope this helps!

Thanks,

[jk615]

i was debating between (C) and (E) but I went with (E) because I thought that was supported by the stimulus. I understand why (C) is correct now, but I was hoping you could explain why (E) is incorrect? Does it describe a flaw, and if so, what would that flaw look like?

[Luke Haqq]

Hi jk615,

For something like (E) to be a flaw, the sentence,

"...we can conclude that most citizens would like to see a legislature that is roughly 40 percent Conservative, 20 percent Moderate, and 40 percent Liberal.

should instead look something like,

• "...we can conclude that exactly 40 percent of citizens would like to see a legislature that is Conservative (or exactly 20 percent would like to see it Moderate, etc.)

That language seems like it would "use evidence that supports only rough estimates to draw a precisely quantified conclusion," as (E) describes.

[pardis]

I cannot understand why it could be concluded that the group wants to see a composition of 40-20-40 based on their responses to the question of which party they would like to see in Congress. I can certainly see why this can't be concluded about individuals but I also can't see how it could be concluded about the group as a whole, as answer choice C suggests.

[James Finch]

Hi Pardis,

The stimulus here presents survey results that show that there was a 40-20-40 split in aggregate of which party the respondents wanted to see in the legislature. It then concludes that each individual person who responded to the survey wants to see a 40-20-40 split in the legislature, as opposed to a Liberal wanting to see a 100% Liberal legislature, or a Conservative wanting to see a 100% Conservative legislature.

This is a Whole-to-Part logical flaw, also known as an error of division. The issue is that just because something is true of a group in aggregate, doesn't mean it is true of each of the group's members. US presidential elections provide a good example of this: in 2016, Hillary Clinton won 48% of the votes for president, while Donald Trump won 46%. If we did what the stimulus here does, we would assume that Clinton won every state by a margin of 48% to 46%. In reality, the electoral college, which is based on the number of states' representatives in Congress, went 307-227 for Trump, because Trump won 30 states while Clinton won only 20, because the proportion of voters who voted for Clinton varied quite a bit state-by-state.

Hope this clears things up!

[pardis]

Hi James, and thanks for your response. I must admit I had to read your analogy several times to overcome my inability to get over the emotions you elicited!!!

I think I didn't clarify the issue I have with the answer choice very well. So let me try again.

(C) states: The argument takes for granted that the preferences of a group as a whole are the preferences of most individuals of the group.

But the stimulus infers the invalid conclusion about individuals within the population, not those within the group.

Is this answer choice still correct because by making the inference about the individuals within the population, the stimulus committed the whole-to-part flaw about the members of the group first?

[Adam Tyson]

Don't forget the part in the stimulus which says "if the survey results are reliable", paradis! The author isn't saying that the population does indeed split along the same lines that the survey respondents did, but rather that IF they do, then most individuals want that split. There's no flaw in assuming that the survey is valid, because his conclusion was conditioned on it being valid. That's like me saying "if my team's quarterback gets injured in the pre-season, then we will have to recruit a new one, because our backups aren't up to the task". I'm not saying that he WILL get hurt, but only what would be true IF he got hurt.

The whole-to-part is the only flaw here that I can see. I hope that helps!

[Moris_cn]

If "most citizens" in the last sentence of the stimulus is changed into "the population", is the argument valid?

[Malila Robinson]

Hi Moris_cn,
Population would still not work because we only know about the population that responded to the survey, not the entire population. (So "population" seems to take "most citizens" and turn it into "all citizens," which doesn't help with the flaw.)
Hope that helps!
-Malila