NS FL 9 CP Q25

Post Reply
mgarc805
Posts: 35
Joined: Tue Jun 25, 2019 11:23 am

NS FL 9 CP Q25

Post by mgarc805 » Thu Aug 08, 2019 11:17 am

In the explanation for the answer it says:
able 1 shows that, over the course of 15 minutes after thawing, retinol levels as measured by fluorometry decreased significantly compared to HPLC
how do you know if that drop if significant I still see a strong correlation with HPLC
NS_Tutor_Mathias
Posts: 264
Joined: Sat Mar 30, 2019 8:39 pm

Re: NS FL 9 CP Q25

Post by NS_Tutor_Mathias » Thu Aug 08, 2019 3:08 pm

So I agree that B is the correct answer here, but you are probably onto something: The use of the word significant in the explanation is not appropriate.

Answer choice A may be tempting, but table 1 does not in itself give us any information about cost and it does suggest that it is not exactly equally as accurate (or more importantly, it also provides no comparison with 15-minute thaw HPLC, which we would need to make the distinction. So even though this is a good inference, we don't have evidence for this answer choice.

Answer choice B uses the correct language for how weak the evidence is: Delays in sample processing MAY give lower values than HPLC. The table does indeed suggest that this MAY be a possibility, therefore this is the only absolutely correct answer choice.

Keep this pattern well in mind when answering MCAT questions: You do not want to just give the most reasonable answer, but preferably one that cannot be refuted - which in experimental design means only drawing the conclusions that can actually be drawn from what has been presented. This goes doubly so for question stems that reference only a single table or figure (or even multiple tables/figures). If the answer over-reaches and requires information outside the table or figure, that is not it.



Beyond that, you are correct. There is no evidence present of a significant difference between the two, even after a 15 minute thaw. Clinical VAD is defined as <100 micrograms per liter and subclinical VAD as <300, and as such the small ~2.5% shift seen by the delay in analysis after unthawing would not hinder a diagnosis at all. Treatment of subclinical VAD is trivial and can be performed at the slightest suspicion, and conditions approaching clinical VAD would be obvious even if your instrument was only +/-10% accurate.
mgarc805
Posts: 35
Joined: Tue Jun 25, 2019 11:23 am

Re: NS FL 9 CP Q25

Post by mgarc805 » Thu Aug 08, 2019 3:38 pm

thank you for your response. I understand how the word "may" can influence the strength of the answer choice and I see how analyzing costs is beyond the scope of the table. but in regards to this question would you be able to assume significance based on R^2 values such as the ones presented in the table. Because I am confused as to how come that they results after 15 min are still highly correlated but that decreased is far enough to assume that it may occur as a result of longer thawing periods.
NS_Tutor_Mathias
Posts: 264
Joined: Sat Mar 30, 2019 8:39 pm

Re: NS FL 9 CP Q25

Post by NS_Tutor_Mathias » Thu Aug 08, 2019 4:21 pm

"May" just means "you didn't prove that this is impossible". In this case we did one better, and had very slim evidence that suggests that our initial hypothesis could be true. You don't have to significance test to say "may". Smoking may make you grow taller. It almost certainly doesn't, but until you sit down and prove that to me, it just might. By contrast, answer choice A makes a definite statement without having the evidence to back it up and is therefore outright wrong.

You wouldn't be asked to do any amount of math here to prove significance on the MCAT in any case, which is why I pointed out that the current explanation provided for the question is flawed and will probably be revised soon. I provided some quick back-of-the-envelope estimates using the clinical guidelines for VAD diagnosis and treatment to show that even a looser use of the term significant would be inappropriate here. But that was largely to keep this discussion grounded in reality and has little to do with MCAT test-taking skills.
mgarc805
Posts: 35
Joined: Tue Jun 25, 2019 11:23 am

Re: NS FL 9 CP Q25

Post by mgarc805 » Mon Aug 12, 2019 11:56 am

Yes I understand but beyond the question or the explanation to the answer, Would you be able to determine significance based solely on R^2.
NS_Tutor_Mathias
Posts: 264
Joined: Sat Mar 30, 2019 8:39 pm

Re: NS FL 9 CP Q25

Post by NS_Tutor_Mathias » Wed Aug 14, 2019 4:57 pm

Significance is ultimately an arbitrary threshold. A discussion of how to create a reasonable cutoff is very, very far beyond the scope of the MCAT or the time and space we have here.
Post Reply