#23 - Several carefully conducted studies

[Adam Tyson]

While that worked in this case, student987, I would say that is a risky approach to take generally in Parallel Flaw questions. In a standard Parallel Reasoning question, we want everything to match, including the number and type of premises and all the conclusions. In Parallel Flaw, however, the primary goal, overriding all others, is that the FLAW be the same. Two arguments that have the same flaw - say, an Error of Composition or a Source Argument - could have different numbers and types of premises, and one may have extraneous information that the other lacks, or one may have an intermediate conclusion and the other does not. Focus on the flaw in the reasoning, and match that up.

It would have been very easy to craft a wrong answer to this question that had an intermediate conclusion, and not that hard to craft a correct answer, with the same numbers/percentages flaw, without one. It's great that you noticed the similarity there, and it's okay to use that to zoom in on an answer and decide to test that one first, but be sure that it does what the question required it to do - contain the same flaw as the stimulus, not necessarily the same non-flaw components.

[student987]

That was so helpful, thank you very much! I was confused about why the way I was solving standard Parallel Reasoning questions didn't work on some Parallel Flaw questions, and it makes much more sense now.

So you're saying that if this were a Parallel Reasoning, the answer choice would have to include an intermediary conclusion, right?

[Adam Tyson]

It should, yes, but remember that the "best answer" never has to be a "perfect" answer, so they could make it harder by leaving that intermediate conclusion out, if that was the best answer in all other respects. Mostly, though, they will have all the same elements in the correct answer as were found in the stimulus.

[student987]

Thank you!