#18 - According to a government official involved in

[PositiveThinker]

I was able to get this answer right simply off of process of elimination.

But How does D weaken? Why must the argument include information about the percentage of all small airplane flights when the pilot whistles? How would that make the conclusion/recommendation that passengers should take safety precautions any less reasonable?

[Dave Killoran]

Hi Positive,

This is an excellent problem for helping you understand one way that numbers and percentages are used on this test, and also how a straight use of a percentage (or number) can be misleading.

First, to help you see how this is working, let me remake the stimulus around an example you've probably seen before:

• According to a government health official, 100 percent of the individuals who died last week drank water during the week. Therefore, if you see a friend drinking water, you should rush them to the hospital, whether or not they appear sick.

Of course, we can see immediately that this is ridiculous, but it shows how the citing of statistics can be entirely useless in certain cases. Every person drinks water with some regularity, so the fact that everyone who died had water intake during the prior week is no surprise. In this case, it's easy to see the flaw above, but the basic operating principle is the same as in our stimulus, it's just that you don't have nearly the same familiarity with pilots, whistling, and crashes as you do with people and water. So, when they change that context, it suddenly seems harder to analyze.

Now, if we look at answer choice (D) in connection to the example above, you can see how it operates more clearly:

• (D) provides no information about the percentage of all individuals who drank water during the week

Of course, the answer to (D) is that everyone (or 100%) of people drank water, which shows that when individuals died that week, the water was unlikely to be the cause, or at least not a signal that death was imminent. In the same way, in answer choice (D) of the problem, maybe it's the case that pilots are just happy people and they whistle a lot. Perhaps if they looked at every small plane flight, they'd find every pilot was whistling, in which case there would be many instances of pilots whistling and no crash followed.

Please let me know if that helps. Thanks!

[LSAT2018]

I understood the error of reasoning for this question, but the wording for answer choice E and D threw me off. For answer choice D, doesn't the percentage of all small airplane flights that involve relatively minor accidents provide more context to support the argument?

[Shannon Parker]

I understood the error of reasoning for this question, but the wording for answer choice E and D threw me off. For answer choice D, doesn't the percentage of all small airplane flights that involve relatively minor accidents provide more context to support the argument?

Answer choice E, the percentage of all small airplane flights that involve relatively minor accidents, has no bearing on the cause of those accidents. It could be 100% of small airplane flights, or 1% of small airplane flights. The statistical error here is that the author uses whistling by the pilot as an indication that there is going to be a minor accident because over 75% of the voice recorders in planes that had minor accidents recorded the pilot whistling. The author does so without any comparison to how many pilots normally whistle. For instance, if pilots whistle in 99% of small airplane flights it would show that there is no correlation.

Hope this clears it up.
Shannon

[LSAT2018]

So essentially this is a correlation/causation problem with the information on 75 percent of voice recorder tapes being used to make a causal statement that, whistling precedes the accident, which pose some safety risk.
Cause → Effect
Whistle → Pose Risk

And the information in the correct answer (D) possibly weakens the conclusion which takes potentially insufficient evidence to arrive at a conclusion. Additionally, answer (E) is incorrect because it does not address the whistling–accident relationship. Would that be the right approach to this question?

[LSAT2018]

First, to help you see how this is working, let me remake the stimulus around an example you've probably seen before:

• According to a government health official, 100 percent of the individuals who died last week drank water during the week. Therefore, if you see a friend drinking water, you should rush them to the hospital, whether or not they appear sick.

Of course, we can see immediately that this is ridiculous, but it shows how the citing of statistics can be entirely useless in certain cases. Every person drinks water with some regularity, so the fact that everyone who died had water intake during the prior week is no surprise. In this case, it's easy to see the flaw above, but the basic operating principle is the same as in our stimulus, it's just that you don't have nearly the same familiarity with pilots, whistling, and crashes as you do with people and water. So, when they change that context, it suddenly seems harder to analyze.

Can I know which practice test this question is from?

[T.B.Justin]

PT 14

[JocelynL]

I'm going through flaw questions and trying to identify the flaw and the way the test makers describe it in the answer choice.
Can someone please clarify which flaw this is describing? At first I thought this flaw concludes a causal relationship when only a correlation is present? Or is this a survey flaw in that only 75% of cases show this relationship and the author applies it to ALL cases? Or overgeneralization?
SO MANY FLAWS!

[Adam Tyson]

You should classify this as a Numbers and Percentages Flaw, JocelynL. The data supports the idea that in 75% of these minor accidents, the pilot was whistling just before the accident. The conclusion seems to be based on the assumption that 75% of the time that a pilot whistles, there will be an accident, which is a misinterpretation of that data. What if pilots whistle all the time, but rarely have an accident? What if whistling pilots almost never have an accident, and those that do represent only a small fraction of all whistling pilots?

Some number to help make it clearer:

100 pilots flew small planes last year.
99 of them whistled during the flight.
4 of them had accidents.
The one pilot that never whistled was in one of those 4 minor accidents.

So, of the planes in those accidents, 75% had whistling pilots, but the vast majority of whistling pilots had no accidents. Do you want the passengers on those planes to all start putting their heads between their knees and assuming the crash position every time Johnny Pilot starts whistling a few bars of the Andy Griffith theme?

It's all about the percentages being misunderstood or misapplied that makes this such a bad argument.