5
$\begingroup$

This answer just consists of a couple of paper references. My initial feeling is that we should treat these like link only answers and ask the author to add more details inline etc so that the answer is valid even if the paper isn't available. Ultimately if this doesn't happen the answers might end up being deleted or converted into comments.

On the other hand academic papers should be more durable/findable in the long-term than random Internet links, so I thought I'd ask a general question here first.

EDIT: the poster has now added the abstracts to the answer, which makes the references a lot more helpful, so for me this question is now just about future cases.

| |
$\endgroup$
  • 1
    $\begingroup$ Some paraphrasing of the source material would also be useful if the source is behind a paywall. $\endgroup$ – LateralFractal Nov 21 '14 at 8:39
  • 1
    $\begingroup$ I think at least a sentence or two about why a reference is relevant should be in order. Papers may be durable, but finding and reading them takes time and there are potentially large efficiency savings if I can determine beforehand how relevant the referenced work is likely to be. $\endgroup$ – Ubiquitous Nov 21 '14 at 8:54
  • 1
    $\begingroup$ Unless the question is asking for references, references don't answer questions. Flag as not an answer. $\endgroup$ – curiousdannii Nov 21 '14 at 10:40
  • 2
    $\begingroup$ I think there are some questions best answered with a reference (not only reference requests), e.g. "Is there someone using method X on the labor market? - Yes, for example paper Y uses this method." It might not be the best answer but imho still a valid answer. (The question you mentioned is imho not answered by the "answer" and not answerable by a reference only, but I wouldn't condemn reference only answers per se.) $\endgroup$ – The Almighty Bob Nov 21 '14 at 12:36
  • $\begingroup$ @TheAlmightyBob: I agree with you, but I still think a few words explaining what is in the references are in order. Remember that the answer is not only for the original asker, but also for anyone in future who comes here with a similar question. That may include people who legitimately can't access the references. It might also include people with slightly different, but related questions who may therefore have a hard time judging whether the references are relevant or not. such people would surely be helped by a brief synopsis. $\endgroup$ – Ubiquitous Nov 23 '14 at 10:32
2
$\begingroup$

Just a few thoughts. Something is always better than nothing and people are giving up their own time to post anything at all. Not that anybody is suggesting it, but I just want to suggest that no one should be punished for posting a link/citation only answer. Hopefully, if the question is a good question, other people will post more complete answers (and get compensated with upvotes or whatever) and the asker doesn't need to accept the answer until he/she is satisfied. So, my recommendation is that

  • citation only answer should NOT be automatically removed
  • don't downvote unless the question was actually harmful (the citation was irrelevant to the question or something)
  • it's ok to ask for more details---always ok. But, we should just remember that people are giving up their own time to provide answers---even if it's just a citation (citations alone can be super useful after all)
| |
$\endgroup$
  • 1
    $\begingroup$ This is not how other SE sites work and I see no reason why it should be how this site works. $\endgroup$ – curiousdannii Nov 23 '14 at 0:09
2
$\begingroup$

Having read the other answers and thought about this a bit more, I don't think reference-only answers should be deleted/converted into comments.

Looking again at Shog9's post about link-only answers, he points out that a link where the link text describes the content - e.g. "You probably want a FileOutputStream" - are valid answers, even if not very good ones.

References also include that kind of information, so we should keep them. They are certainly better - i.e. more worthy of upvotes - if they include a synopsis and explanation of why they are relevant, though.

| |
$\endgroup$
1
$\begingroup$

A link-only answer is problematic when the user then has to follow the link to find the right answer, for the following reasons:

  1. A lot of people browsing might be unable, and many would be unwilling, to follow a link away.

  2. We're trying to build a body of awesome content here. While it's great to link to awesome content elsewhere, it's even better to summarise it here too: that way, economics.SE can become a known substantial repository for expert knowledge.

  3. When new people do happen by, typically from a web search, we'd like them to stay and contribute quality content. If, when they arrive for the first time, all we do is send them elsewhere to find their answer, our rate of accumulation of new members of the community will suffer.

  4. Summarising the link adds indexable content that search engines can use to help other people with similar questions to find our content here.

So, we should discourage link-only answers, with the usual tools: commenting to give guidance on best practice; down-voting; and flagging as "not an answer".

As an additional note while we're on the subject, it's best if journal links are via DOI (using one of the major DOI linkers such as dx.doi.org) whenever possible. There's a good chance that that will be more stable in the long term than some deep link to a particular publisher or journal. And it would make it easier for SE devs to do a search & replace in the databaes, if dx.doi.org ever did go down.

| |
$\endgroup$
1
$\begingroup$

I'm the OP in that question and I would actually be satisfied with references only if the papers were relevant, which they were not.

I'm asking whether anyone has dealt with argument XYZ, so if I get some references that deal with it, I'm happy. If I want to be sure that its the right paper, I can still look into the abstract.

The link-only policy on Stackexchange was also established because of the dependency on the external website - if the link breaks, the answer becomes useless.

However, with scientific references (especially published ones), there is no way the "link breaks" - so this point is somewhat invalid.

Ah, and regarding paywalls: Most authors have the final working paper version of the publication online, so there's always a "free" version available. I have not encountered else, at least.

Update

Actually, the (lengthy) descriptions that it contains now are copies of the abstracts. I feel uncomfortable with that, not only due to potential copyright/plagiarism issues.

| |
$\endgroup$
  • 2
    $\begingroup$ Abstracts are pretty well always freely available online. I can't imagine how that could possibly be a breach of copyright or plagiarism, particularly if it comes with a link to the journal or database. $\endgroup$ – Jamzy Nov 25 '14 at 1:54

You must log in to answer this question.

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