12

I am interested in learning what users of this site look for and appreciate in an answer.

  • On this site, what are the characteristics of a good answer? (Examples of exceptionally good answers, and what makes them so exceptional, are welcome!)

  • What are the characteristics of a bad answer on this site? (if such a thing even exists...)

Possible aspects to consider include but are not limited to: length, content, style, tone, disclosing background of answer-er or not, citing outside sources, addressing question in general vs situation-specific way, answers from users with or without specific relevant experience, bias, etc.

Of course, not all questions benefit from the same kinds of answers, and responses addressing subtleties like this are also very welcome.


(This post is shamelessly stolen from Biblical Hermeneutics meta. Yes, I read other SE sites' metas, including those of sites I don't participate in at all.)

  • 6
    "Yes, I read other SE sites' metas, including those of sites I don't participate in at all." I approve of the level of SE nerdiness at display here :) – xLeitix Jan 21 '15 at 7:15
  • 2
    Personally, I usually upvote almost every answer I read that isn't terrible. – Ben Bitdiddle Jan 21 '15 at 20:40
16

What I look for in answers: (i.e., what I typically upvote)

  • A neutral, down-to-earth tone
  • A fresh take on a question (i.e., don't make your answer start with "I agree with XY")
  • Substantial answers (very short answers are not typically very useful to me)
  • A user that, based on her/his bio and SE habitus, seems trustworthy to answer the question
  • Sources, if appropriate for the question

The last-but-one bullet likely requires more explanation: given that we usually deal with rather subjective topics, I usually take into account who posts an answer if it goes against my own opinion or seems counter-intuitive. Yes, that's unfair towards new users, but I have certainly seen new users post, well, stupid things much more frequently than high-rep users, who, often, also happen to be senior academics.

What I hate in answers: (i.e., what I typically downvote)

  • Agenda answers ("this clearly shows that [professors|students|...] are just ..." - we had a few of those recently)
  • Answers that seem to fall into the "uninformed opinion" category ("I don't have experience with this, but clearly ...")
  • Excessive strong language, uncalled-for attacks towards the OP or somebody else mentioned in the question
  • Circumventing the actual question asked, and instead answering what the answerer thinks "should" be the question (although there is definitely a substantial grey area here)

Bonus point: What I hate in questions:

  • More than everything, I hate questions where the OP has clearly a pre-formed opinion, argues with everybody who answers differently, and then goes ahead to accept the first answer that validates her/his opinion no matter how much more votes all the other answers had.
  • 3
    Another question hate: I hate shopping questions. – jakebeal Jan 22 '15 at 4:04
  • 1
    Another reason: the answer is to the point and addresses what is asked; for the same reason, I upvoted this answer. – OK- Jan 22 '15 at 9:54
  • Along the line of your last paragraphs, There are users who accept answers that the community clearly indicated were wrong or sketchy. – Cape Code Jan 22 '15 at 20:06
  • @CapeCode The second one seems particularly interesting as there were a bunch of highly-voted better answers. This is kind of what I hate - you don't necessarily need to accept the highest-voted answer, but if you don't care about what the community thinks at all, then why ask? – xLeitix Jan 23 '15 at 4:17
9

I don't know if this question asks about what we should look for or what we are looking for. I ran this query and extracted the answers that received 100 votes or more*.

Here are the links to the best voted answers to date in decreasing order of vote count:

232 - 182 - 156 - 154 - 152 - 151 - 147 - 128 - 120 - 118 - 113 - 106 - 102 - 101 - 100

Some observations (listed in the order used by xLeitix in his answer):

  1. Most have a polite, down-to-earth tone. One ('Mind your own business') has had its tone disputed and one politely questions the good faith of the original question.

  2. None of them are 'support' answers repeating or confirming an existing answer. One is even clearly at odd with the others.

  3. The average answer length on Academia.SE according to this query is 1212 characters. The character count in the listed answers (obtained with a Firefox plugin) are: 376 2234 5964 1098 2354 3389 6303 3091 3476 1692 2535 793 3686 2100 279 in the respective order. Mean: 2624.66; standard deviation: 1791.99; and median: 2354. They are substantially longer than average, with a few exceptions.

  4. High-rep users are well represented: 3 are from JeffE, 9 out of 15 are from 20k+ users. But there are exceptions and some are even almost the only answer given by the poster. It seems like we are still pretty open to newcomers inputs.

  5. Only one cites a reference document (a policy description on a university website), 2 give links to Wikipedia articles, and one to a book. None cite scientific publications or data. Citing sources does not seem to be a criteria for success.

* I know it's more a measure of popularity than quality, but I think it still shows which answers we are looking for. There is an obvious bias due to the popularity of the question, two questions have actually multiple answers in this list. Popular questions attract popular answers, and it's not a surprise.

  • 1
    I love that answer. re: the 'Mind your own business' one - I think this is an artefact more than anything else. This style of answer would likely not go down well with many questions, but apparently many users felt that it was appropriate for this one. Also, Jeff has a bit of a trademark here for very direct statements and answers. I am not sure if the same answer given by @userXXXXXX would have received as many upvotes. – xLeitix Jan 22 '15 at 7:07
7

Adding to the other things already said, I also think it is very valuable when an answer delves into the principles and reasoning the lead the poster to answer in the way that they did. I think that this is particularly valuable because many answers are derived from a broader scientific or pedagogical ethos. Communicating that ethos helps beyond the specific situation in the question, and helps to build and reinforce the better elements of academia.

  • 1
    Good point. This is the "show your work" principle. It's a common value across many SE sites (some require it more strictly than others). – ff524 Jan 29 '15 at 1:50
1

xLeitix's answer is great, just want to add one thing to that: answers should be concise. Many answers here tend to have lots of examples or stories or other stuff that's relevant but not required, and it almost always makes the answers much harder to read with little benefit. Shorter is almost always better.

  • 5
    I actually disagree with that. I think examples etc. very often improve answers substantially. – xLeitix Jan 23 '15 at 4:18
  • There is a balance point... explanation is good, multi-page walls of text like this answer not so much. – jakebeal Jan 23 '15 at 15:16
  • 3
    @jakebeal Well, I think it goes without saying that multiple pages of incoherent ramblings are not what we are looking for in an answer. – xLeitix Jan 25 '15 at 15:12

You must log in to answer this question.

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