4
$\begingroup$

I've noticed a bunch of disputed not-an-answer flags in the queue recently. While it's great that people are actively flagging things, this trend does suggest some disagreement on or misunderstanding of what these flags should be used for. Is that really the case? Are people clear on the purpose of the not-an-answer flag, and if not, is there anything we should be doing to make it more clear?

As a first step, let's try something. All of the following were flagged as not an answer by multiple high-rep users, and the flag was disputed, often by multiple high-rep users. What I'd like to hear from you:

  • If you would flag any of these as not-an-answer, what's your justification for doing so?
  • If you would dispute an existing not-an-answer flag on any of these, what's your justification for doing so?

Ideally, I'd like to get at least one opinion on each side for each answer. The point is not to correctly guess whether each of these is NAA, so even if (especially if) your justification turns out to be "wrong" I still want to hear it. The point is for us moderators to understand whether, and how, people are misunderstanding the purpose of the NAA flag.

On this question:

Your question should more accurately have been, "Why does sound travel faster in solid iron than in liquid mercury even though mercury has higher density?"

Were the question phrased that way, the answer would be more obvious. At temperatures at which both metals are liquid or both metals are solid, sound travels faster in the denser metal.

and on this question:

The chemical bonds in the dye would be broken by high energy UV photons from the sun - the moon has no atmosphere or ozone layer to protect it.

Why so cheap? Remember NASA in the 60s was a different organisation, led by an engineering challenge to get to the moon and back, nothing else mattered. Need a flag? Just get whatever fits.

A more modern Nasa would have had an entire division of flag-ness populated by PR professionals and managers. With nationwide competitions for k12 children, a careful check that the flag supplier employed a suitable number of disabled veterans and each star made by a different agency in each state. All resulting in a $5,000,000 flag with it's own twitter following.

Of course it wouldn't have any scientists and engineers to actually get it to the moon.

and on this question:

When you say the black hole would need to be incredibly huge for this to be possible, are you thinking of it's size from the perspective of our universe? This seems wrong to me because the black hole would have to be located in another much much larger universe than our own. So even though from our perspective it would seem huge, in the universe where the black hole exists it could be average sized. Kind of like m-theory but instead of equal sized branes an infinite progression of smaller and smaller (or bigger and bigger) universes.

and on this question:

The birds must feel something, because they usually sit on the ground wire, not the high voltage wires.

Nothing special about these answers; they're simply the latest ones. But I've been seeing this pattern for quite a while.

$\endgroup$
4
$\begingroup$

I just looked at my flagging history and I did indeed go through those four examples. I flagged two of them as Not An Answer, and I disputed the other two.

I flagged

  • the moon flag one, because it is more of a rant than anything, and most of all it does not address the core of the question, as to how bleaching might be avoided, and

  • the observable-universe one, because it is confusing, doesn't really say much, and simply piles on speculation on an already large helping.

I disputed the other two. Looking at them again, I guess they do fall on the wrong side of the line. I'm not very sure why I flagged like that, to be honest; it may partly have been a desire to not be too harsh on new users.

  • The sound-speed-in-mercury one is, I think, not that far from an (incorrect) attempt at an answer. It happens fairly often that someone takes a bad question and reformulates so that it has an answer; this looks like a failed attempt at that. I still see it in a bit of a gray area.

  • With the current-through-birds one, I felt if they could back up their assertions it might eventually become a full answer, even if incorrect. I guess I also acted partly to balance out the other comments, which seemed out of proportion.


To answer your more general question:

not very much.

Flags are hard to handle. In particular, it has happened to me a few times that I see a very bad answer and flag it, and it gets declined with the message

declined - flags should not be used to indicate technical inaccuracies, or an altogether wrong answer.

These were on answers with very limited value to anyone (example, example). If it turns out that the Very Low Quality flag does not cover those answers, what does it cover? This sort of thing has made me quite jumpy at the time of flagging posts, and it tends to make me err on the side of forgiveness rather than harshness. I note now that this was, I think, only for the VLQ queue and not for the Not An Answer one, but I can't really separate the two in my mind / recollections all that clearly.

$\endgroup$
  • $\begingroup$ Thanks, this is some of the kind of feedback I'm looking for. So perhaps we'll make some attempt to have publicize an accessible explanation of what these flags are for. (TL;DR: the intent of the SE team, as I understand it, is that VLQ is for posts that are garbled or unparseable as English text, and NAA is for posts that are a separate question or a comment, or otherwise make no attempt whatsoever to answer the question.) $\endgroup$ – David Z Mar 5 '14 at 21:55
4
$\begingroup$

Disputed:

Sound speed in mercury: I felt this correctly pointed out (albeit in a roundabout way) that apples were being compared to oranges, and so the density-sound speed relation shouldn't even be expected to hold. Sure it might be incomplete, but the OP didn't give any specific direction for ways to go, so I can't say this isn't an answer. I feel answers that could be better are still answers.

Flagged:

Universe as a black hole: Seemed like it was asking for clarification, and as such was a comment, not an answer. I tried to come to this decision without worrying about the correctness/non-mainstreamness of the answer.1

Birds on a power line: This didn't answer the question asked. If the OP had asked "Can birds detect electricity flowing through power lines?" then this would be an acceptable answer. Again, it may be factually wrong, but it answers the question in that case. However, the OP asked for a quantitative result, and this only gives a qualitative answer. In that way it is a comment, not an answer. A qualitative answer that explains why a quantitative result would be nigh impossible would be acceptable, but this post doesn't support itself at all.

Untouched:

Flag on the Moon: I agree it's an out-of-place rant that went a bit too long. Mitigating circumstance: The OP asked for it ("Why so cheap? Didn't NASA know...?"). Also, if left by itself, the first sentence would count as some sort of answer (at about the same level of completeness as the sound speed case above). Question: Can a post with an answer be deemed "not an answer" due to other stuff in it?


As for the overall purpose of NAA: I feel I know how I use it, in the sense that I have self-consistent criteria that I apply objectively. And I feel my flagging states suggest I'm mostly in line with the moderators and broader community on this: 159/176 (90%) NAA/invalid/moderator attention flags deemed helpful. And my own feeling is that the flags I see from others aren't far off from what I would judge myself. The first page of my flagging history has 46 NAA's I've issued and only 5 invalids. So again, ~90% agreement.2


1 If "not an answer" applied to posts that were completely wrong, based on a grave misunderstanding of basic principles, well then we should flag most of the posts on this site that use the word "horizon." But that's an exasperation for another time.

2 Caveat: I will sometimes be on the border and ignore something in the flag queue until it goes away. So there is missing data in a sense.

$\endgroup$
  • $\begingroup$ "Can a post with an answer be deemed "not an answer" due to other stuff in it?" I would say no. That sounds more like a case for either editing or downvoting, depending on the nature and extent of the non-answering material. $\endgroup$ – David Z Mar 8 '14 at 1:00
2
$\begingroup$

NOTE: When I write up the first part of this, I am referring only to the context provided at the questions and answers themselves, at the time I am writing v1 of this answer. In the second part of this answer, I will address other comments provided here.

Iron/Mercury sound: I would flag as Not an Answer. That answer just says, "If you asked this slightly different question, then the answer would be trivial." I would be okay with that, if the answer provided an explanation of why the original question doesn't make sense, and why that re-interpretation of the question is natural or needed. I would also be okay with that if, in addition to that, also answered the question as actually asked (in fact, I frequently do this myself). But the answer doesn't provide any of that.

Flag on Moon: I would not flag. I would dispute an NaA flag. The third paragraph is not helpful, and should not be there. By itself, it would be Not an Answer. But the rest of the answer actually does address the question.

Universe in black hole: I would not flag. I would dispute an NaA flag. I read this answer as an attempt to interpret the question so it makes sense, and answer what the answerer regards as the question. I contrast this with the Iron/Mercury question because the answer here explains why he thinks he needs to apply his understanding to the question.

Birds on a wire: I would not flag. Without the comments provided there by Emilio and Brandon, I would dispute a flag. The comments point out the the premise of the answer is incorrect or needs to be justified. If the premise of the answer is incorrect, I regard it as Not an Answer (though I don't always flag such answers). I would not dispute an NaA flag if I saw either of those comments attached to the flag.


My reactions to other answers here:

Question: Can a post with an answer be deemed "not an answer" due to other stuff in it?

I want to say that, in theory, an answer could contain enough off-topic content for me to regard it as Not an Answer. In that case, the off-topic content would have to overwhelm the on-topic content to such an extent that it becomes difficult to find the on-topic answer. I don't see that in this case. I'm sure I've seen that happen in the past, but I can't think of any examples. UPDATE: Ron Maimon's answer to a GR resource request question is a perfect example of what I mean by this. The first sentence should just be a comment, and the rest of the answer doesn't even attempt to answer the question. I have just flagged it as Not An Answer.

If "not an answer" applied to posts that were completely wrong, based on a grave misunderstanding of basic principles [...]

and

it has happened to me a few times that I see a very bad answer and flag it, and it gets declined with the message

declined - flags should not be used to indicate technical inaccuracies, or an altogether wrong answer

In the past, I have flagged things that I found so egregiously wrong or incoherent from a physics standpoint that I regarded them as not answers. And I have see that reply from whoever answered my flag. The result is that I am slower to flag things as NaA than I used to be. The answers I gave above about what I would have flagged (not necessarily about what I would have disputed) are, I believe, in line with what the team and the mods intend for that flag to mean.

NAA is for posts that are a separate question or a comment, or otherwise make no attempt whatsoever to answer the question.

I think there are two points where I disagree with what I think is the official intent of the NAA flag. One is the Birds on a wire answer.I believe the intent of the flag is not to include that answer as NaA, and leave it to voting to sort out the value of that answer. But I don't consider that an attempt to answer the question unless/until the author adds a justification for the premise of that answer. And I think we should consider deleting some answers like that if they don't add that justification after some time. The second point is that some answers are attempts to answer the question in a way, but the actual physics content is just incoherent. These answers are not just wrong, they're Not Even Wrong. Again, my understanding is that the official intent is to downvote those into oblivion, but I think we should consider deleting content that is physically incoherent.

$\endgroup$
  • 3
    $\begingroup$ We have a problem with site size when it comes to deleting poor answers. We have only five non-mod users with the vote-to-delete privilege. Flagging as VLQ will put a question in the VLQ queue where 2k+ users can Recommend Deletion (6 votes delete an answer). However, this is hard to reconcile with the 'declined' message you quote. $\endgroup$ – Emilio Pisanty Mar 6 '14 at 11:11
  • 1
    $\begingroup$ I hope the answer by Ron Maimon that you linked to doesn't get deleted. It's of questionable relevance to the question, but it's a lovely piece of writing and it'd be a shame to see it go. $\endgroup$ – Nathaniel Mar 17 '14 at 3:07
  • $\begingroup$ @Nathaniel, "It's of questionable relevance to the question," is why it should be deleted. It doesn't matter how "lovely" it is; it doesn't belong there. You're an 11k-rep user; you should understand this. $\endgroup$ – Colin McFaul Mar 17 '14 at 12:26
  • 1
    $\begingroup$ Understand yes, agree not really. A hard-line insistence on rules over judgement is not something I regard as fruitful in this community. (And it is fairly specific to physics.SE in comparison to the other more specialised stack exchange sites.) That post has value; there is no tangible benefit in destroying it. $\endgroup$ – Nathaniel Mar 17 '14 at 13:13
  • $\begingroup$ @Nathaniel, that post has no value. This is not a hard-line insistence on rules on my part; I am valuing good judgment here. If you think the post has value, then copy and paste it somewhere where it's useful. $\endgroup$ – Colin McFaul Mar 17 '14 at 14:11
  • 1
    $\begingroup$ Earlier you seemed to concede that it was indeed "lovely", but you claimed it had to be deleted because, apparently, it's not enough for its relevance to be questionable, it has to be absolute. Now you've abandoned that argument in favour of claiming that your judgement is objectively better than mine. I am, in fact, wrong in claiming that it has value to me, because your X-ray specs tell you it has value to no-one. This is exactly the sort if attitude that makes me question my participation in this community. $\endgroup$ – Nathaniel Mar 17 '14 at 14:23
  • $\begingroup$ @Nathaniel, I actually didn't concede that it was "lovely;" I don't find that relevant at all. And I actually don't find this answer to have "questionable" relevance (as you concede); it's irrelevance is absolute. On this issue, yes, my judgment is objectively better than yours. That's because I can see that this answer is irrelevant to the question it's "answering," and you can't. That's what makes my judgment here better than yours. I don't doubt that you find it useful; if that's the case, then copy and paste somewhere you can keep it. $\endgroup$ – Colin McFaul Mar 17 '14 at 14:40
  • $\begingroup$ And if you think the answer is relevant, then defend that opinion. Otherwise, there's nothing further to discuss. $\endgroup$ – Colin McFaul Mar 17 '14 at 14:41
  • 1
    $\begingroup$ All I said was I quite like it as a piece of writing, and that I would be sad to see it go. That is what you have chosen to argue against. I do not have to defend it as being relevant, because the idea that it can only have value if it has "absolute" relevance is entirely your own invention. "Has value" does not mean the same thing as "satisfies a pre-defined set of objective criteria". If it did then it would not be a value judgement. $\endgroup$ – Nathaniel Mar 17 '14 at 15:24
-1
$\begingroup$

Just a quick note on a relevant development at MSO:

Let's get rid of the 10K flag queue

Essentially, if it goes through, much of this discussion would be obsolete. I quite like it, by the way.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .