Q10 - Columnist: The dangers of mountain climbing

[ohthatpatrick]

Question Type:
Flaw (technically, Weaken, since prefaced by 'fails to consider')

Stimulus Breakdown:
Conclusion: The popular media greatly exaggerates the dangers of mountain climbing (i.e. Mountain climbing's not that dangerous)

Evidence: There were only 200 deaths on Everest over 80 years, whereas there were 7000 traffic fatalities in France in just one year.

Answer Anticipation:
This is a percent vs. number flaw. To analyze the riskiness of something, we want to know what percent/proportion of the time something bad happens. Raw numbers aren't very informative, because without knowing the size of the underlying population, we can't judge how scary something is / isn't. Suppose that last year only 200 people got sick by eating Tide Pods while 500 people got sick by eating shellfish. Sounds like shellfish is scarier right? What if I say, "out of the 1000 people who ate Tide Pods, 200 got sick. Out of the 1 million people who ate shellfish, 500 people got sick." That changes our risk perception.

Similarly, maybe only 1000 people try to climb Everest, whereas millions of people drive in France.

Correct Answer:
D

Answer Choice Analysis:
(A) Sure, this weakens a tad by suggesting that maybe 2002 is an anomalously high year for the traffic fatalities stat. But this is really just saying, "what if 2002 isn't the LOWEST ever recorded". If it's not the lowest ever recorded, then by definition the number in other years would be lower.

(B) This is the same as (A). Yes, it's possible the author cherry picked France (or 2002), but there are so many countries and so many years that even if France is higher than most, it still doesn't seem like an unreasonable number to choose. What we need is not about France vs. other countries, but "number of people driving in France vs. number of people attempting to scale Everest".

(C) This would only strengthen the author's position that mountain climbing isn't as dangerous as some say. We could make it even safer by enacting certain measures.

(D) YES, and man this answer choice is one of those classic, "If you don't already get it, this answer won't help you get it" type answers. Given that danger would be measured in terms of PERCENTAGE of people who die, not NUMBER of people who die, the author needs to consider the underlying totals of French drivers and Everest climbers.

(E) This is similar to (A) and (B). Maybe the author cherry picked an unflattering year (2002) or country (France) for traffic fatalities. Maybe the author cherry picked a flattering mountain (Everest) for climbing fatalities. All these answers would cancel each other out. And again, our primary concern with the current reasoning is that even if Everest and 2002 and France are fair things to look at, we still can't judge the danger without knowing the underlying populations.

Takeaway/Pattern: This is an interesting set of answer choices, since I would say the original argument is definitely committing % vs. # as well as Sampling flaws. But since the 7,000 is 35 times greater than the 200, even if 2002, France, or Everest aren't the most representative picks, they'd have to be MASSIVELY unrepresentative in order to bring the 7,000 and the 200 closer to each other. Meanwhile, sampling issue or not, we would always have a % vs. # issue in judging the raw numbers as any metric of "danger/risk".

#officialexplanation