kungfury42 wrote:
GMATNinja I would really appreciate if you could please help here. I think option E should be correct instead of option D. Here's why:
Quote:
D) In Pelmenia, women have a significantly lower life expectancy than men, and women are also more likely to be born left-handed.
This option doesn't tell us anything about the sex ratio of Pelmenia. Suppose there are 5 women born for every 1000 men in Pelmenia. And suppose out of every 5 women born, 4 are born left-handed and out of every 1000 men born, 200 are born left-handed.
Now we can see that women are 80% (4/5) likely to be born left-handed and men are 20% (200/1000) likely to be born left-handed which is consistent with the passage that women are more likely to be born left-handed than men.
Now as we can see that since women are significantly fewer in number than men, so even if women have significantly lower life expectancy than men, it wouldn't make any difference in the %age of people who are left-handed as we progress on the age. This eliminates option D.
Quote:
E) As Pelmenians age, they become less likely to participate in surveys regarding handedness or hand dominance.
I want to quote the famous coin-toss example here. Suppose if we toss a coin infinite times, say 100,000,000,000 times we will have a 50%-50% occurence of head and tail in our sample set.
Suppose if we reduce the number of coin tosses to 100,000 we will still have something like 52%-48% occurence of head and tail in our sample set.
But if we reduce the number of tosses to a very small number say 100, we will have nowhere near 50%-50% occurence of head and tail. Maybe we will have 60 Heads and 40 Tails or 75 Heads and 25 Tails.
This is to say that as the sample space reduces, we cannot expect the reduced sample space to follow the same %age pattern as the original sample space. However, as we keep increasing the sample space, the distribution is bound to converge to a value (50-50 in this case)
This serves as a clear evidence that as Pelmenians age and become less participative in surveys regarding handedness we are bound to get skewed results in line with what's shown in the stem.
We cannot assume that a sample space of 100 Pelmenians (at the age of 70) will share the same distribution of handedness as do the 10,000,000 Pelminians (at the age of 18).
Hence, we can say that option E is the correct reason that explains why percentage of left-handed Pelmenians is getting significantly changed as we move up the age groups.
Posted from my mobile device Let's start with (E):
Quote:
E) As Pelmenians age, they become less likely to participate in surveys regarding handedness or hand dominance.
In the passage, the data skews in a certain direction -- the percentage of
left-handed people goes down dramatically.
(E) talks about Pelmenians in general, including both right and left-handed folks. As these Pelmenians get older, they are ALL less likely to participate in surveys. This doesn't explain the specific trend in the data that the passage describes -- why are the left-handed people
in particular disappearing? (E) really doesn't give us a reason for this phenomenon.
To your point about sample size: sure, the size of your surveyed population does impact how much you should trust that data! However, (E) really doesn't give us enough to say that the sample of 45 year-olds and 70 year-olds is small enough to throw out the results. We have no idea whether the sample size went from a million to nine hundred thousand, or from a million to ten.
So,
maybe if you squint sideways at it you could say that (E) implies that some of the data isn't perfectly reliable -- but it gives us nothing to explain why the percentage of left-handed people in particular is diminishing.
(Also, as a side note: the coin example is a bit off! If you flipped a coin 100,000 times, you'd expect to get heads on about 49.5% to 50.5% of those flips in over 99% of your trials. And if you flip a coin 100 times, a 75/25 split would be an extremely unlikely outlier too -- you might want to get your coin checked for fairness
. So, while sample size definitely matters, it doesn't have as big of an effect as you might think).
Compare that to (D):
Quote:
D) In Pelmenia, women have a significantly lower life expectancy than men, and women are also more likely to be born left-handed.
(D) tells us that our left-handers are dying off more quickly than our right-handers, and that gives us a great explanation for why the percentage of left-handedness reduces with age.
The edge-case that you've come up with is quite extreme -- you'd really need more information to assume something so outlandish as a skewed brith ratio like the one you're describing. But even in your scenario, if the women are dying earlier than the men, you'd expect the percentage of left-handed people in the population to go down as the population ages. So, (D) would help explain the data.
And, even if you've come up with one crazy scenario that makes (D) a bit weaker, that's fine! We're looking for the answer choice that "most helps" to explain the phenomenon, so we don't need a
perfect answer choice.
(D) gives us a reason that the incidence of left-handedness goes down with age, while (E) does not. So, (D) is our answer.
I hope that helps!
close
Updated on: 02 Nov 2018, 06:00
43%
(01:45)
wrong 
