PT60, S1, Q9 - In one study of a particular plant species

[chlqusghtk]

Would you please explain why (a) and (c) cannot resolve the paradox here? I got this question right and understand why (d) is the right one, but curious what the reasons for crossing out (a) and (c) are.

Usually, I'm not very comfortable with this kind of questions--stimulus containing numbers and percentages. When a question asks to weaken the argument with some percentages(e.g., PT 59, S3 Q13), how could you know when an answer choice talking about numbers can attack the argument and when it cannot?

Thanks a lot!

[giladedelman]

Thanks for posting!

So, the first study found that 70% of a particular plant species had patterned stems, but the second study, of the same area, found that only 40% did. What gives?

Well, if the first study used a broader definition of "patterned," then we would expect it to include more plants under that definition. So (D) makes sense.

(A) is irrelevant, because the stimulus is about the percentage of plants that have patterned stems. The overall plant population wouldn't be expected to affect the ratio of patterned to unpatterned stems.

(B) is out of scope. We're talking about "a particular plant species."

(C) is tempting, but why would including more plants be likely to lead to a different result? Again, we're interested in the percentage of plants with patterned stems. It's possible that the first study had too small a sample to be statistically significant, but that's too big a leap to make from what the stimulus gives us.

(E) is irrelevant, because we're told what we're told. It doesn't matter whether the data was collected as a primary or secondary goal.

Does that answer your question on this one?

Now, as for percent problems in general, I'm not sure if I can give any specific advice. Maybe something to keep in mind is that percents refer to proportions of groups, not absolute numbers of things.

Why don't you post a question about the other problem you mentioned?

[chlqusghtk]

Thanks for the explanation!

But what still bothers me is that... what if you have the same number of patterned stems but different overall numbers of stems?

That is, if both of the studies found the same number of patterned stems--the number that becomes numerator, but they differ in the total number, which would be denominator, wouldn't they get different percentages? Even if they have different numerators to begin with, when their denominators get changed, the resulting percentages would be also changed, no?

That's why I was confused about (a) and (c); both of them are talking about changes in the denominator part.


Would you please help me clarify this?


Thanks!

[giladedelman]

What you say is true; if you found the same number of patterned stems, but more stems overall, then you'd be dealing with a lower percentage of patterned stems.

But hang on -- where do answers (A) and (C) say anything about the number of patterned stems (the numerator, as you correctly put it)? They don't! They tell us that the denominator is bigger, but that doesn't mean a darn thing about the numerator! So you're making a huge leap to try to justify these answer choices.

Imagine I conduct a study of 20,000 dolphins, and find that 50% are female. That means out of my sample, there are 10,000 female dolphins.

Now let's say I do a second study, this time of 40,000 dolphins. Do we still expect to find 10,000 females? No! As we increase the sample size, there's no reason to expect the percentage to change. We'd expect to find 20,000 females -- 50%.

Similarly, we would expect the two plant studies to turn up the same percentage of patterned stems, regardless of the size of the sample.

Does that make sense?

[chlqusghtk]

Yes! Yes! It perfectly does.

So, what you are saying is that since both (a) and (d) only talk about denominator, not numerator, they do not provide enough concrete information to solve the discrepancy b/w the two percentages. Am I correct?

Thank you so much for your help!