Crowdsourcing in a nutshell

One of the things we are promised the Web will do for us is answer all our questions without anybody having to be paid to collect and collate the information. Whatever you ask, someone will know the answer.

Only an out-of-date cynic would say that a random person with no known credentials may not be the best source of information.

What this out-of-date cynic did was look at Yahoo! Answers. I wasn’t looking round for a target to croticize: I genuinely wanted to know the atmospheric pressure at 3000 metres, because I’ve skied at that height and I’ve been able to breathe.

Here was the question, from ‘TimR’. It looks like a homework question to me.

  1. The rate of change of atmospheric pressure P with respect to altitude h is proportional to P, provided that the temperature is constant.
  2. At 15 degrees celsius the pressure is 101.3 kPa at sea level …
  3. … and 87.14 kPa at h=1000. [presumably, 1000 metres]
  4. (a) What is the atmospheric pressure at an altitude of 3000m? [presumably, at 15°C, though the question doesn’t say].
  5. (b) What is the pressure at the top ot Mount McKinley, at an altitude of 6187? [presumably, 6187 metres; presumably, at 15°C although that is impossible at the top of the mountain].

Before we start, a little observation. The pressure gets less with height. This is general knowledge anyway, and just in case you didn’t know, 2 and 3 give you two heights to be getting going with.

Here is the Yahoo! Answers answer, from ‘Steve’:

Since they are proportional, we can say that P=h (in terms of there relationship to one another–as one increases, so does the other, and vice versa. This concept is very important to know and allows further understanding of what is meant by “proportional”).

Ignore the 15 degress C part, because you don’t need it to answer this.

(a)
Set up a ratio.

(Pressure1)/(Altitude1) = (Pressure2)/(Altitude2)
It doesn’t matter which pressure/atmosphere you designate as 1 or 2. Just be consistent.
x/3000=87.14/1000
Solve for x; this will give you the atmospheric pressure at 3000m.

(b)
Same as (a), but instead of using 3000, use 6187.

This answer is “Best Answer – Chosen by Voters”, with a rating of 100%.

“Since they are proportional, we can say that P=h” – no, the question does not say that the pressure is proportional to the height.

“as one increases, so does the other” – no, pressure does not increase when height increases. It decreases.

“Set up a ratio. (etc)” – This is wonderful, and it should be carved in letters of gold above the door of physics labs.

  • It says that the pressure at 3000 metres is three times the pressure at 1000 metres. This is false, but uninterestingly so.
  • It says that the pressure at sea level is zero. If this were true, we would all be walking around in space suits. It also happens to contradict what the question explicitly said, which is that the pressure at sea level is 1013 millibars.
  • My favourite, though, is that it says that on the shores of the Dead Sea the pressure is negative. I don’t know what a negative pressure could even mean, but at least now I can travel to the Dead Sea to find out.

Yahoo! Answers is the Web at its best: questions asked, answers given by a panel of clearly expert answerers (Steve has given 52 answers, of which 23% are “Best Answer”), answers checked and rated by other people – and yet we get this. Because even in the new wonderful web world, if getting it right is nobody’s job, nobody has a reason to get it right.

What applies to Yahoo! Answers applies to all the major support forums: Apple’s, Adobe’s, Microsoft’s. What a support ‘forum’ amounts to is this: your question is answered by someone who knows less than you and hasn’t read your question anyway. In a well-attended forum you don’t just get one such answer, you get many, all wrong in different ways. The only way a forum can work is if, in the end, there is a guarantee that someone who knows will look at the forum, edit out (or at least flag) misinformation, and if at all possible give correct answers.

Because in the end truth and falsehood need actual thought. They can’t be determined by statistics.

By the way, if you have come here looking for the answer, it is this. The question says that the rate at which the pressure declines with height is proportional to the pressure. The question says that the pressure at 1000m is 0.86(ish) times the pressure at sea level. So the pressure at 2000m will be 0.86 times the pressure at 1000m and the pressure at 3000m will be 0.86 times the pressure at 2000m. That gives the first answer. The reason for the second part of the question is that the first part can be answered by successive multiplication but the second part needs the concept of raising to a non-integer power: in this case, the power 6.187.

Advertisements