What do we want Academia.SE to be?

Question

Edit: This was not intended to be a "Choose #1 or #2 proposal." This was intended to be a prompt for discussion about what the current community wants from this site. #1 and #2 are just things I've heard from users in the recent past. I am looking for answers that address the general idea of "What should Academia.SE be?" whether they relate to these themes or not.

Recently, we've had several disagreements on meta that I think stem from slightly conflicting views of what this community is (or should be). Broadly, these are:

1. We are building up a library of concise, clear and correct questions and answers as an archive for interested readers. The main goal of this site is to help both those asking questions and future Google users. We want to make it as easy as possible for others to get/find accurate, focused answers to questions and problems that are within the scope of this site.

2. We are facilitating "on-the-record" problem-solving between academics/researchers with questions and academics/researchers with answers. The goal of this site is mostly to connect users with questions to other users who want to answer those questions. The main purpose of a thread is to help the people who participated in it, with the entire thread preserved for the benefit of participants and future viewers.

Both of the sites I've just described can be extremely useful and valuable resources (I think Cross Validated is an example of the former, and MathOverflow is an example of the latter), but trying to be both at once causes some friction. For example,

• Editing: The first Academia.SE encourages editing - improving question titles, removing "noise" ("sorry for the dumb question," extended thanks, over-lengthy and irrelevant personal information, and pleas of desperation) while still preserving the intent of the OP. The second Academia.SE has a hands-off approach to other people's content, in which edits are discouraged unless a post is made CW to explicitly signal that edits are welcome.
• Closing questions: On the first Academia.SE, questions that are outside the scope of the site (as defined on meta) are "noise" and should be closed and/or edited to fit the scope of the site. On the second Academia.SE, if a question is broadly relevant to academia and someone is willing to answer it, it should stay open/be reopened if it's closed.
• Comments: To the first Academia.SE, comments that no longer add any value to the author of a post or to future readers are distracting, and should be removed (once they no longer serve a useful purpose). To the second Academia.SE, comments are a part of the historical record of the communication, and should preserved (except for offensive comments).
• Answers: On the first Academia.SE, answers that might be useful to the OP but are not really answers to the question should be downvoted, converted to comment, or deleted, so as not to distract from "real" answers. On the second Academia.SE, answers that aren't really answers are useful content, and should be treated as such.

I don't think either of these approaches is especially bad - but trying to be both at once, or to be different things to different users, is (I think) not healthy. As a moderator, it makes it difficult to act on the community's wishes (which part of the community should I follow in acting on flags)? It is confusing to new users, since the site policies are so inconsistent (both in action, and as expressed on meta). And I think it leads to conflict between users (in a bad way, not in a healthy way).

We have grown quite a lot (and graduated!) in the last year, and I think it's well past time to revisit what this community is and what primary role it serves. Do we want to tend more towards Academia.SE #1 or Academia.SE #2? How far do we want to go in whatever direction we choose? What goals are most important to us, as a community?

---

Note: this is not a question about a specific scenario or a specific site policy. This is a general question about the future direction of the site, please answer accordingly.

Answer 1

I think that the question presents a false dichotomy. We do not have to choose one option vs the other. Also, as users or moderators we are exactly like comments. We are ephemeral (as moderators tend to remind us about comments). Who knows which one of us will still be around to answer questions or moderate this forum in ten years from now? Why do we need to strictly enforce one policy or the other, based on our "vision" we have for this site. Why do we need to actually have a vision for this site, instead of letting it evolve naturally? Why can't we decide on a question-per-question basis or on comment-per-comment basis? Why do we actually need to have a prevailing policy after all? We are mostly scientists and there maybe more than one correct and viable approaches to the exact same problem exist. Why must we enforce the prevailing one? Academia SE really does not have to be our little science project where we calibrate the results to best fit our theory.

The SE policies have been tested multiple times on many sites and they really work. They are harsh but they help avoid the disintegration of SE sites into a mash of newbies who ask silly questions and chase away the serious users who really want to contribute. But here, we do not have any of those problems. We do not have newbies who want to solve their homework. We do not have "Java HasMap. Plz hlp" questions. We do not have help vampires who answer their questions to increase their rep. The high-rep users are established people who just want to contribute and would never waste their time to answer idiotic questions for increasing their rep. As such, this community has none of the problems of Stack Overflow as pointed in many meta questions there (e.g., here and here).

Also, on a personal view, I am not very interested in rounding different questions and answers so that all the different use cases are rounded to the exact same problem, to be marked as duplicates to facilitate easier Google searching. 90% of the answers in this forum may be answered by canned answers like, "Ask your advisor", "Do not walk, run", "Do research as an undergraduate and get good grades to get a good PHD position", "You are not too old for a PHD" and so-on. Do we really need rounded / identical questions and canned answers or we need personalized, high-quality advice from Academia professionals? This is the real question and what we should decide after all. And to that end, option 2 is a a far better choice.

Answer 2

Honestly, I think that we can and should effectively serve both goals.

I see the aim of this site as consistent with #1: providing consistent, clear, and correct answers.

Because of the nature of the questions and answers, however, there is inherently more subjectivity and personal perspective involved in creating a good answer than there is for a question on programming or statistics. For this same reason, I believe that in many cases even answering a fairly question can provide a much more general value in the answer: it is for this reason that I typically take pains to explain the principles and reasoning process that I use to arrive at an answer, because I think that in many cases those are more of the takeaway than the answer itself. I see many others doing the same, and suspect it may be for the same reasons.

From these basic principles, I would propose the following approaches to the various policy questions:

• Editing: Removing "noise" is fine, but should be careful to preserve as much of the original intent and also voice of the OP as possible.

• Closing questions: Off-topic questions should be closed, but when a question is borderline it is best if it closes more slowly, rather than getting a mod-jump to the end. I see a lot of questions that get a quick answer and/or de facto answers in comments that probably adequately address the OP's problem, even though the question is then deemed not worth of retention. I think this is a good balance between problem solving and curation.

• Comments: Many comments serve as good "footnotes" to the original posts, as well as information about how an edge case is best resolved. We should definitely remove obsolete comments, but be slow about it and generous about what we consider a footnote.

• Answers: Again, I think we should be fairly generous in what we consider to be an on-topic answer, because many times the exact question asked is not actually the question the person actually wants answered, often due to some degree of ignorance on their part. Anything that doesn't fit a fairly generous reading, however, should be downvoted, converted, or deleted.

As you may notice, I think that in principle we should hew to #1, but in practice due to the nature of the comments there will be a lot of #2 as part of the reality of providing good answers to many issues, especially around judgement and ethics.

Answer 3

tl;dr: vigorous maintenance and weeding help make this site the success it is

Stack Exchange software and guidelines are all designed for "building up a library of concise, clear and correct questions and answers".

That's not the only way to do things. As userxxxx noted in a now-deleted answer, places like the Ask Academia sub-Reddit provide a great venue for other approaches.

But expert Q&A about Academia is what we do well - better than anywhere else on the web, IME. It's how we do things on Stack Exchange - that's why most of us are here rather than somewhere else. There are other places for long discussions all over the web, and have been for many years. There are other places for opinion pieces, for mutual support groups, and so on. We don't need to try to provide those platforms as well as being a Q&A site, because there are already plenty of places on the web for those, with software platforms tailored for them.

The software platform here is tailored for expert Q&A. That's what we've been so far, that's what we're good at: we've got a great web niche, and we fill it really well.

We don't have help vampires, precisely because we are a type #1 site: we close and delete nonsense quickly and efficiently.

We have the tools to close duplicates; to close opinion-based questions; to keep discussions contained to chat. And we use them. Academia.SE has one of the lowest ratios of questions in limbo (closed, but not deleted) of the whole network.

I agree with Alexandros that we're not here to maintain a tiny number of canonical questions and close everything else as duplicates. At the same time, having a hundred answers scattered across fifty questions that all boil down to variants of the same answer is an unmaintanable mess, and that would get in the way of us helping each other.

We've had the discussion about comments here several times, and every time we come back to yes, comments are designed to be ephemeral, and that's what works.

There's a lot of experience across over a hundred Stack Exchange sites about what works in the short run and in the long run. Those are embodied in the guidelines that are common to almost all the sites. They all point to building a library of concise, clear questions and answers. Not discussion. So that means deleting comments once they've served their purposes of seeking clarification or pointing out egregious errors. That means closing questions that are not clear, are off-topic, are too broad, that are primarily opinion based; and then getting those closed questions either edited and reopened, or deleted.

This is about whether we continue with the meticulous weeding and maintenance tasks, or let the site get overgrown with whatever happens to self-sow here. There are already plenty of briar patches all over the internet, and it would be a real waste to take on of the web's best-maintained academic gardens and let it decay into just another briar patch.

Answer 4

Academia in generally is pretty harsh and we all have a lot of things to rant about. What I really like about AC.SE is that we are NOT a discussion board and we have a low tolerance for ranting. Many of our questions are "soft" and our answers are our opinions driven by our experience. While our answers may be our opinions, I think sometimes others can better express MY opinion better than I can. I would like to see AC.SE as a community driven site where we all work together to improve the questions and answers by freely editing them to be as helpful and clear as possible. I think that adding an answer that simply tweaks the meaning of another answer just adds noise to the site. We need to be able to edit things. If answers and comments are not directly relevant, they should be improved or deleted.

I think closing questions is the biggest issue. Keeping AC.SE rant free is important to me. Questions should be closed if they are outside our scope and most definitely if they are discussion oriented. It doesn't matter if someone wants to answer the question or not.

Answer 5

It depends on who you want to get as targetted audience, both as active contributors to the site AND as "passive" external readers.

If the goal is, as the name Academia of the site implies, to have a high expert-level site useful to and used by academics such as professors, active researchers, posdocs, ect you have to treat these folks and their generous contributions with the appropriate respect and appreciation. The schedule of academics and researchers is often very time constrained, which means that the time and effort they voluntarily offer to contribute to something such as this site, can not be taken for granted. If they dont get something back from contributing to Academia themself and dont think the site is of use and helpful to them, they simply dont come or leave again.

So if the targetted audience of Academia is academics and active researchers, the only appropriate (and demonstrably successful) way to moderate the site is the MathOverflow model (Option 2), which is in addition to MathOverflow also successfully applied (adopted exactly) by the research-level Theoretical Computer Science site in the SE network.

From an academics/active researchers point of view, option 1 is orthogonal to any high-level academic community, it just wants to lure academics and experts in to write a dead online book useful for an as large as possible general audience of random external viewers for free, without giving anything back to them or allowing the site to be useful to the people who actually offer their time and effort to write the content.

In summary: if a high density of accomplished academics and active researchers is important for Academia, the SO model (option 1) is not suitable and the MathOverflow model (option 2) should be adopted.