33

I have recently noticed a trend in the answers on academia.se to question the situation about which OP is asking because not every single detail is provided, instead of answering what is being asked.

Most recent examples from a few days ago include:

Should professors intervene if a student is wearing offensive clothing in their classroom?

  • The question asks what to do about clearly offensive clothing worn to class as a TA or other person of authority
  • Some answers, instead of answering, argue that the OP is easily offended, or that what was offending to her is not offending in general

Is it ethical to apply different criteria for graduate admissions based on country of undergraduate study?

  • The question is asking if it is ethical to design admission criteria according to a certain statistics their research revealed (it is a bit of a controversial admission criteria). (example)
  • Instead of answering, some answers were suggesting that the statistics and the research they did must have been wrong and the results are not valid. (example)

These are just two more recent examples. I know when I was asking a question lately, I was sure to cover all my bases (e.g. explicating that everything was done ethically and in good will) to avoid answers questioning my motives and methods, since I saw this kind of non-answers prevail and even be the most-upvoted answers often.

I understand that questions asking clearly unethical things, such as How do I best cheat on the admission process, or something similar, should not be answered, but this is not the case I am talking about.

One other thing is that people say "if you do not want to disclose the details, there is something wrong or unethical in your methodology, because you don't want to disclose the details." I understand the need for anonymity, or the wish to generalize, so I don't think this is a good trend

While some of the opinions might be valid in case the OP really made a mistake, those kind of answers still do not answer the question actually asked, and it is not our place to question the facts presented.

Yet, the community seems not to be condemning those kind of answers very strong: neither of the examples I linked to has negative score (one has a positive one!), even thou, (and, please, correct me if I'm wrong), they do not actually answer the question.

Do we really want to collect such answers?

And, as a secondary questions, what do we want to do with such answers that already exist? Should we flag-delete them? We could wait for the down-votes to push them sufficiently down, but especially on a bit controversial questions, those answers seem to get support from part of the community because of their attitude towards the controversial issue in question, and not because they actually offer an answer.

  • 2
    I noticed the problem on the offensive-clothing question and upvoted your question but I am not sure the statistics-by-origin question is really similar. It's difficult to separate the validity of the approach and the ethics of using it. If you really think there is a major flaw, how could it be ethical to select students based on that? And how could you seriously discuss the ethics aspect while glossing over the issue? – Relaxed Mar 7 '14 at 6:50
  • 1
    I agree with you, this is something I've noticed as well in various threads around here (can't give specific examples right now). It often happens that instead of taking the information in the question as an assumption and answering accordingly, people are suggesting their own interpretations of the situation, which contradict the interpretation given by the OP. (EDIT: I've only now realised it's an old thread. It was linked to in the sidebar...) – Pandora Jul 19 '16 at 16:01
21

A fundamental rule of flags is: flags are NOT to be used to delete incorrect or bad answers.

That is, if there is an actual attempt to engage the question, you cannot ask the mods to delete the question just because you think it is inaccurate, gives bad advice, or challenges the assumptions in the question instead of accepting them as fact.

The correct way to express your displeasure with such a question is to downvote it.

  • 1
    I guess you meant "displeasure with such an answer". I know about the flagging "rule", but these cases seem borderline, why I asked on meta. An exaggerated example would be a question "As a PhD student, I'm facing such and such problem", somebody answered "I don't believe you're a PhD so I'm gonna answer as if you were a permanent" which is clearly not an answer. If somebody says "it was clearly offensive" and somebody answers with "I think it was only offensive to you, and in such case you should...", it is not so clear. I could probably find more such question-answer examples than two above. – penelope Mar 6 '14 at 13:51
  • @penelope on the two questions referenced, the posts that really did not address the question at all were (appropriately) flagged as "not an answer" and deleted. The other "bad" answers in those examples at least address some part of the question. – ff524 Mar 7 '14 at 6:20
9

As one of my answers was pulled out as an example, I think I should chime in here. Let me start by saying I think this is an important question and needs to be discussed. Further, to be clear, I do not feel attacked, singled out, or defensive. I think that the highlighted answer is a good example of the issue.

The issues with these types of answers is that the person writing the answer may be convinced that they are not denying the situation and think they are providing a helpful answer. The votes on answers like these are not particularly meaningful. As you say,

those answers seem to get support from part of the community because of their attitude towards the controversial issue in question

but I would also claim they get down votes because of the controversial issue in the question. Flagging the question doesn't generally work since most answerers will not see the flag. It also doesn't provide a place for discussion and puts mods/high rep users in a difficult situation. It seems to me that the comments to the answer and/or chat (and possibly meta) is the place to discuss and figure out what is going on. It might be informative to go back to my answer to see how I thought about it.

The title of the question is:

Is it ethical to apply different criteria for graduate admissions based on country of undergraduate study?

The only statement within the body of the original question (and current version) with a question mark is

Is it fair to apply different criteria to students from different countries in admissions decisions?

I think that the number of up votes on the question suggest it is an important question for our site. Despite my up vote for the question, I don't think the question is a particularly good fit for the SE format since the question "is X ethical/fair" is essentially a yes/no question and providing an evidence based answer is difficult. I think this is confirmed by the number and variety of answers as well as the up votes and down votes of the answers.

Moving on to my answer and the "charge" levied against it:

Instead of answering, some answers were suggesting that the statistics and the research they did must have been wrong and the results are not valid.

We need to first decide if I answered the question. My first sentence could (and probably should) be reworded to be

I would argue that you are using the results in an unethical and discriminatory way because you are interpreting your data incorrectly

That seems to me to be a pretty clear answering of the question. I answered "No, it is not ethical". I then tried to provide reasoning for why I answered the way I did. My answer is not particularly great in that there is little evidence to support my claim. I don't want to clarify/defend my answer here as it is too far removed from the answer itself, but I am happy to continue to clarify/defend my answer in either the comments to the answer or in chat. At the heart of my reasoning is that based on the original question, the edits to the question, the comments, and my chat discussion with the OP suggests that the classic mistake of interpreting correlation as causation is being made. I tried very hard in my answer to not claim that the statistics were wrong, but purely focused on the interpretation. In summary, I disagree with both claims about my answer (not answering the question and denying the situation).

Currently the answer has 3 down votes and three people made "negative" comments prior to the last down vote being cast. From the negative comments it is clear that my answer is confusing to people and could use an edit. The up votes (and one positive comment) suggest to me that some people see the value in my answer. Based on the mixed feedback, I would normally edit my answer to try and improve it. In this case, I feel it is better to leave it be at least for a while so that this discussion about the issue of answers that miss/deny the point can be addressed.

  • 2
    I agree that your answer did address the question, since it went on to say that you considered it unethical to use the results of the study, and so it would not have been appropriate to flag+delete it. (An edit would help clarify that you aren't denying the situation of the question, since as currently written it does give that impression.) – ff524 Mar 7 '14 at 19:00
6

I think commenting and downvoting are probably the best options here. I suspect flags will be ignored for the reasons @aeismail mentions. The problem is that the "community" (defined as readers who vote) appear to have conflicting views on the matter, and what you're asking about is a form of minority protection that SE doesn't really have a mechanism for.

But I know that I read comments very carefully, and that I'd be influenced by comments pointing out that the answer is not answering the question as stated. I tend to be less likely to downvote (and I'm not sure why), and maybe that's part of the problem.

5

The behaviours you are seeing of questioning the question is due to the enormous prevalence of X-Y questions in the original StackOverflow and other STEM StackExchange sites. Questions where the thread is "how do I do X", but experienced users see that the OP has a more fundamental misunderstanding, and should really be doing Y. The SO model is so popular because follow-up comments let us determine the true nature of the OPs intent, get to the bottom of the question, and answers can either target the question-as-asked, or the OPs fundamental issue if it deviates. This is a good thing. Answering all questions at face-value is not helpful -- otherwise why was the OJ Simpson thread closed?

  • And yet that attitude can lead to what I see as a pervasive problem on SO/SE, which I think this question relates to a specific example of: somebody asks "Given X, how do I Y?" and gets a load of responses saying "Don't do X" - often in a belittling way. While on some occasions this may be constructive, as you note, often it is based on assumptions from a limited understanding of the original problem and is unhelpful, while coming across as arrogant and aggressive. I know a number of people who've been driven away from SE by such responses, which they see as "nerds showing off". – Flyto Apr 22 '16 at 13:02
  • I can appreciate that, but I think the alternative is not no arrogance but rather no nerds. Some of the most helpful places on the net, mainly IRC, is also some of the most hostile to new comers asking how to do X and getting criticized for ever wanting to do X in the first place. I literally can't count the number of times i've thought to myself "god, what a jackass this guy is, telling me to choose a different database schema when all I need to do is query X. So high and mighty..." -- only to find out either seconds or days later that I really do need a different schema, not a better query. – Wetlab Walter Apr 22 '16 at 15:34
  • In the cases where i've wanted to do something unusual and actually the recommended solution was wrong (for example, I often use exec() in my python code for performance reasons, and i'm constantly called out for it) -- usually I just have to explain why i'm being a special flower, and everyone benefits. I agree it can be frustrating, but there is no utility in answering bad questions without mentioning there's a better way -- which usually requires more information to determine. – Wetlab Walter Apr 22 '16 at 15:36
  • You see "hostile yet helpful", many just see hostile, aggressive, arrogant, entitled, etc. What one person (most often white and male) sees as "tough love", another may see as "you have no right to be here, you are too ignorant and/or Not Like Us". – Flyto Apr 22 '16 at 15:46
  • 2
    I have no idea what white-male, entitlement, or human-rights have to do with this discussion at all. All i'm saying is that in the vast majority of circumstances where something feels amiss - something is amiss. – Wetlab Walter Apr 22 '16 at 15:48
3

Answers questioning the question are ultimately non-constructive.

Such things should be dealt with on of the following, instead of posting answer:

  • comments under question if someone asks for more details,
  • down votes,
  • close votes.

One of powers of StackExchange is that is is not a forum.

However, I remember some questions when the correct answer was that someone is interpreting situation incorrectly. (But it is rather an exception than rule.)

And what we should do with such answers?

  • If it is "not even an answer" then should comment suggesting posting it as a comment to question (and delete the answer).
  • If it has parts of an answer (just plays down the importance of issue, or anything) - do as with any other answer we think that adds negative net value: downvote (it's better to stay democratic here).

(A separate issue, perhaps worth a different question on meta, is whenever to provide direct examples. In some cases, for example the "clearly offensive clothing" it might cut some idle discussion (though, I would advice to use different example of a similar calibre, for the sake of anonymousness). For example in question What to do if assignment is against student's religion? an example helps in avoiding "guessing game".)

  • Judging from the question about the offensive material, it seems like providing an example would not be possible, as any example of a similar calibre would likely be edited out (see the answer by HdS). Though of course, one could then add one and edit it out and point to the edit history for those who want to see it. – Tobias Kildetoft Mar 6 '14 at 11:40
  • 1
    @TobiasKildetoft In such case I prefer concrete texts, even if written in a descriptive way (in above, "equates one of mainstream religious figures with a vulgar word for private parts"). For me, word offensive is very vague (and sometimes it is related to one's taste, in other - it is clear violation of law; different cases may allow different actions; e.g. from wearing military-related outfit, to being dressed in full Nazi uniform). In any case, it is material for a new question on meta. Or maybe even two. – Piotr Migdal Mar 6 '14 at 12:27
  • 1
    That is actually a very good way to provide a more concrete example without presenting possibly offensive material yourself. – Tobias Kildetoft Mar 6 '14 at 12:28
  • Well, I think the OPs sentence "The item of clothing in question contained a slogan and image that is indubitably demeaning and hostile towards women." defines the problem fairly well. But, while I agree with what you said, it still doesn't suggest how to deal with such answers that are already posted. Should we flag-delete them, or is something else appropriate? – penelope Mar 6 '14 at 12:29
  • @penelope Edited (for "what to do with such answers"). From the mentioned question it is way better than general "offensive t-shirt", but still, gives me a lot of room for interpretation (and I as saw, for other as well). Only after reading further comments, I realized that it is rather serious thing. Things like, for example, "states that women should not pursue professional career" or "advertises physical violence" would, IMHO, shorten discussion. – Piotr Migdal Mar 6 '14 at 12:48
  • 5
    I felt that details would only engender further discussion and doubt: for example, "states that women should not pursue professional career" -> "Maybe the shirt only said something like 'I love my stay-at-home mom'?" There really is no end to this kind of thing... – ff524 Mar 7 '14 at 6:17
  • 3
    I think my classification of the slogan as "indubitably demeaning and hostile towards women" was specific enough to answer the question, and that more information would not be constructive. – ff524 Mar 7 '14 at 6:18
  • 1
    @ff524 Maybe you are right and more information wouldn't help (and even would stir even more discussion whether something is "offensive enough"). In any case, it is different topic, so I move it to: meta.academia.stackexchange.com/questions/842/…. – Piotr Migdal Mar 7 '14 at 18:35
1

I think you have to be careful to not eliminate the best possible response to some questions. Many times the best response to a question is simply not an answer (or JUST an answer). For example, if someone asks for the best way for a high school freshman to break into math research, the best response must at least include some sort of serious challenge to the OP (whether or not it gives the most plausible way for the freshman to break into research). This is an extreme example, but in general many responses should challenge the assumptions of the OP, while also giving an answer. Because these are often human matters that are extremely emotionally charged, it is important to challenge the OP if they are in false dichotomous thinking, for example. But in almost every case I think this should be accompanied by an ANSWER, GIVEN THE ASSUMPTIONS OF THE OP.

0

IMO, this is a species of a more general problem on discussion forums. Members too often want to contribute when they don't have anything to contribute. Those members often write posts that either,

  • tell the OP that they should not give too much thought to their concern,
  • tell the OP that they should go about things an entirely different way that happens to render the question moot,
  • tell the OP that they're wrong about a subjective matter, or
  • question the OP's reasons for wanting to do what they want to do.

In any case, these replies suck. We should ban them.

You must log in to answer this question.

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