How closely does the LIS Stack Exchange need to follow other Stack Exchanges in terms of "appropriate" questions?

Would it be feasible to open the LIS Stack Exchange site up to the types of questions that are not allowed on other Stack Exchange sites? I'm trying to figure out how to make this site more useful to a professional librarian audience. Right now, it seems like the questions that are most easily accepted are relevant to library users, not the experts that Stack Exchanges tend to try to encourage.

Many of the questions I have don't have the "one true answer". For instance, I'd love to know things like, "We're planning to migrate to a new ILS in 2014, what are you using and what do you like or dislike about it?" or "What companies would you recommend to provide analysis of your patron base?" or "What's a good way to solicit community feedback for features in our next ILS?" When I do have a yes/no question I either use one of the more generic Stack Exchanges or post to my ILS vendor's listserv.

It's not just me, the Library Loon, notes that:

Does LIS have such [clear-cut] problems? Of a certainty—common cataloging and metadata problems, for example, or some reference stumpers. But this is a fairly impoverished set of problems, hardly the sum total of the difficulties professionals encounter in LIS. The problems that fall outside that circumscribed area are… messy. Complicated. Fuzzy. Not always susceptible to obvious or easily-evaluable solution. Lacking the One True Answer that StackExchans engineered to elicit.

Shopping questions are very useful in the library world. Even some of the questions that get through could be considered "shopping questions" versus objective questions. Take Holiday Decorations in the Library, it's a tricky, yet necessary, question that gets debated on Library listserves every Christmas. However, in essence, it's a shopping question. Why should I "buy" this library policy over that library policy?

Gem

• discussion

Comments

• jonsca: I'm not sure we agree on the definition of a shopping recommendation. To me, a shopping rec is something that is asking for a product or service (free or not). I understand your complaint with the other policies, but is there a way those questions could be reformulated to make them more "canonical" (which would be okay in my book) rather than a survey of "what does your library do?"?
• jonsca: (I do, by the way, genuinely appreciate you stepping up with your opinion on the matter, as I think we need to hear more from our userbase before we make any decisions.)
• Gem: Hm, I'm not sure I understand why there's a difference between a product/service and anything else that doesn't have one right answer. There are so many variables with even the simplest of questions in libraries. That even a policy that works for one library wouldn't for another, similar to a software product.
• Gem: Would it be valid to ask for what the downsides of a specific product are? I would consider that a shopping question. However, leaving it free-form allows me to learn of downsides I would never have considered before. Such as "integrated" according to one ILS vendor means you have to have two different windows up and duplicate your checkouts in each window. Not what most of us would consider integrated but the vendor does. Libraries change ILS products every 5-10 years (at max) so we don't have enough in-house comparisons to catch some of these things that other libraries have experienced.
• 1. Alan Thomas II: There are no "canonical" answers not derived from case studies when it comes to lots of things about LIS; that's one of the fundamental points we're trying to make here. (See my answer below for a fuller discussion of that.) However, there are major differences in the arguments for and against different types of question sharing the same general nature of "banned on SE in general for reasons that don't necessarily apply to us." Shog9 seems to think that we should break each of those out into different questions (i.e., he seems willing to have the discussions), so I'm going to try that.

Answer by Shog9

Just a quick note on Shopping questions...

This term gets over-used on a lot of sites, to damn perfectly good questions. A shopping question is problematic when they approach a problem with a solution already in hand (a product to buy) and ask someone else to... Well, "go shopping" for them.

What’s the point of a bunch of labor intensive questions that provide only temporary benefit to a limit (some might say Too Localized) audience?

They're bad, they waste time, and you should do was much as possible to divert the folks asking them in a better direction.

But not every question looking for a product or service is a "shopping" question!

Ask yourself this: can I use my knowledge and experience to provide (or verify) a solution to a real problem faced by the asker, one that might possibly include the mention of one or more things they could purchase in order to meet their needs, without having to regularly visit the shops or websites of various merchants in order to verify that my answer continues to reflect market conditions?

Then it's probably not a shopping question. Now, maybe there are other things wrong with it, but forcing answerers to go shopping isn't one of them.

In the example you gave, I don't see any sign that the author wants us to go shopping for her. Presumably, she's perfectly capable of tracking down deals on whatever crepe-paper-and-cardboard adornments her organization requires - she just wants advice on what sort of things she should be shopping for!

And that is a perfectly sensible thing to ask.

Comments

• 1. Alan Thomas II: Why is having a standard in hand a problem in our discipline rather than a way of avoiding a subjective argument over what the standard should be, given that there's not always an objectively correct standard? Why do you think that I'm capable of knowing what the available products in my discipline are or are like and need a standard to judge them by rather than vice-versa? Why do you think that a blog post based on the short-product-cycle world of technology applies to all products in our discipline? Why do you think that quoting a blog post and linking to it twice makes it applicable to us?
• 1. Alan Thomas II: Also: Why do people keep linking me to things (sometimes twice in one answer!) and quoting them to me and expecting me to suddenly be enlightened by them as if I must never have been exposed to them before because all it takes is exposure to convince me of their truth? Believe it or not, some of us already read the blog and are not convinced that its posts always accurately address our discipline; that is rather the point of the question that you're failing to answer. Knowledgeably apply the quoted post to our profession per my comment above and I'll reconsider my stance on it.
• Shog9: I said nothing about standards. And frankly, I'm not convinced you read anything else in this answer either. I suggest you start with the second paragraph, first sentence, and then explain how either of those comments has any connection to what I actually wrote.
• 1. Alan Thomas II: You're right, we're talking past each other. Let me back up and address that first: Did the questioner use the wrong term? Yes, he's asking about more than just shopping questions. Did his example not fit the term? Yes, it wasn't a shopping question. But your answer attacks his example rather than answering his question, which makes it better suited to a comment (and, indeed, the issue was already being addressed in the comments).
• 1. Alan Thomas II: The problem that he's asking for solutions to is that the general SE standards posted places like the blog do not always seem to work with or apply to our discipline; your answer relies on a blog post that doesn't always apply to our discipline. I was trying to ask you questions about how the blog post applied to us to show how your answer could itself be seen as an example of the problem, thereby illustrating the underlying problem and hopefully provoking a response that applied to it.
• 1. Alan Thomas II: If you'd like the line-by-line: "A shopping question is problematic when they approach a problem with a solution already in hand (a product to buy) and ask someone else to... Well, 'go shopping' for them." —> "Why is having a standard in hand a problem in our discipline rather than a way of avoiding a subjective argument over what the standard should be, given that there's not always an objectively correct standard? Why do you think that I'm capable of knowing what the available products in my discipline aror are like and need a standard to judge them by rather than vice-versa?"
• 1. Alan Thomas II: (The blog post suggests asking for criteria—a standard—rather than a product, when the problem we actually need solved is frequently the opposite.) "What’s the point of a bunch of labor intensive questions that provide only temporary benefit to a limited (some might say Too Localized) audience?" —> "Why do you think that a blog post based on the short-product-cycle world of technology applies to all products in our discipline?" (The argument for being Too Localized is that the answer won't be relevant in a t is that always true for our discipline?)
• 1. Alan Thomas II: YES, I sometimes need someone to go shopping for me! Many library solutions are very one-shot due to the implementation and transition costs, so we have to poll people who made different choices to find out what the results were. Many vendors are AMAZINGLY opaque about what their product actually does or even looks like, and trials are time-consuming and fail to expose many problems. The product cycles can also be far longer than for, e.g., a camera. In short, the blog post you authoritatively cited has very little to do with us; THAT is the problem we want you to address.
• Shog9: The blog post I linked to is pretty clear about the sort questions it applies to, and the potential problems with them. If you're not talking about that sort of question, and those problems aren't a concern, then using them as a reference is pointless... And yet, here you are apparently trying to argue that even though Gem's question *wasn't* a shopping question and *doesn't* exhibit these problems, that such questions are worth chasing after? Fine - bring one up for discussion then, preferably as a separate meta question. Burning four comments to reiterate what I already wrote is a waste.
• Shog9: BTW: no one ever argued that shopping questions weren't useful to the person *asking* them. The question you need to ask (yourself, and the rest of the folks here) is: are these questions we want to spend time *answering*.
• 1. Alan Thomas II: My general feeling is that at this point we're not answering almost no questions at all because no-one is asking any, so I'd rather open a variety of non-standard categories that people might find useful to questions that meet certain standards and use our community moderation powers to close the questions that don't meet those standards than worry about taxing our community with having to answer >1 question per day.

Answer by M. Alan Thomas II

This is a very open question in that there's a lot of different types of questions which may require different arguments; I'm going to address three such categories of question in separate sections below.

On reference, readers' advisory, &c.:

It has always been held that this site is NOT for the answering of reference questions, RA, &c. That is, even if the question is one asked of a librarian and only answerable by crowd-sourcing our collective expert knowledge and wisdom, if it is not itself about LIS, it's not allowed. That's not a bad point; this is an LIS site. On the other hand, some questions really do require librarians to answer ("name that book" springs to mind), and as a site for experts to ask each other the kind of questions that need other experts to answer, I think that we could have a discussion about these questions. Personally, I'd rather try letting the wisdom of the crowd moderate bad questions than refuse an entire category without the experiment and suffer for it.

There is a second line of attack that says that these questions will flood the site with non-expert questions if we let them in. I disagree. If we do get questions from non-experts that are best served on other sites, we can direct them there (and if that site is an SE site, questions can be moved directly), but if it's from an expert, I'm betting that they're not going to ask a question of us that it would be faster to find out on their own, especially if they've got a patron waiting on it. Even if it is by an expert but could be answered by a different community, we can still redirect them. As things currently stand, we get so few questions that redirecting a few in order to claim others that are relevant isn't going to tax us much; if that changes, we can re-visit the policy at that time.

There is a third line of attack that readers' advisory is subjective, which I feel is slightly insulting to the people who do it and teach it; while it's not perfectly accurate, and therefore not 100% objective, if we couldn't actually predict readers' likely preferences in reality, there wouldn't be any reason to try or anything to teach about the subject, which just isn't true. This line of attack is simply untrue.

I would rather try community moderation to select for good questions than ban entire categories up front, IFF we may be the best community to answer them.

On shopping questions:

The original blog post against shopping questions was oriented around a tech recommendation that would be outdated in a year. The fact that such a small window of relevance made the question Too Localized was bolded to emphasize that it was the real sticking point with the question. For comparison, RPG.SE decided things the other way, I'm guessing in part because even an old edition of an RPG isn't necessarily superseded by later editions for all purposes and in any case the product cycle for a specific product line may be five years, ten years, twenty years, or even infinitely long (i.e., there's only one edition ever). Given that context, it's no wonder that they bucked the trend. What's the situation with library shopping questions? I don't know. It might depend on the question; the ebooks situation is evolving full-steam, but I don't know that that's true for all possible examples.

The other point about shopping questions that was made in the original blog post was they they wanted the answer to be a universal learning experience by discussing what criteria made a good X rather than which products met that criteria at any given time. However, the similar questions that I've asked here have been ones for which I knew which criteria were good ahead of time, I just had no way of researching possible answers that might meet those criteria without crowd-sourcing possibilities. In the engineering and technology world, it's quite possible that asking people what my standard should be is an objective question because mechanical operations have relatively deterministic results, but asking such a question in LIS is easily to open a can of worms you could fill entire journals with (and people frequently do). Similarly, the specs for a wide range of tech items are a quick Google search away and all I need is a filter, whereas I might know what I want for a library item but can't actually find a list of such products anywhere online, much less detailed descriptions of them. Thus, what aspect of the shopping question needs a crowd-sourced answer is very different for us than for the sites that blog post was written about.

I would support shopping questions where the questioner can identify a reasonably objective standard for what they need and the answers are unlikely to change significantly in the next several years (as determined by the wisdom of the crowd using Close —> Too Localized votes).

On polls, best practices, and subjectivity:

As has en stated, we are not in a discipline with a lot of provably correct answers, at least outside of some technical areas (physical questions such as book preservation, many information science questions, &c.). We ARE in a discipline in which "How we done it good" (as one of my professors put it) is a valid and common type of professional journal or magazine article. In short, we are in a discipline in which polling the community for how they have tackled a problem, discussing the answers, and then collectively selecting one or more of the answers as "best" is how we arrive at canonical best practices. In SE terms, these steps are called "asking a question," "answering a question," "comments," "voting on answers," and "accepting a best answer." They're also called "subjective" and "non-constructive."

Libraries.SE is in a position to not just report canonical answers that everyone already knows and agrees on but to discover for the first time what the canonical answers are. That is, I think, something that we should embrace. No, we can't write a bright line rule for determining in advance which questions are going to discover a consensus best answer, because we won't know if there's one to discover until the question has been asked and people have started talking about it. Yes, this might be seen as subjective and non-constructive by the standards written for sites where "Does it compile?" is a reasonable test for whether something is a valid answer (whether or not it's correct). However, it's how things work in this discipline and it's constructive in the best possible sense: it can discover hidden knowledge and construct a database of it. Isn't that what we're trying to do here? If SE does not want to support that form of inquiry, SE does not want to support an LIS site, and we need to know that now so that we can go elsewhere.

If a question just wants to know "How do YOU do things?" it's a bad question, but if it wants to know "What's a good way of doing X?" it can be constructive if there's some obvious goal that the answers have to be working towards (i.e., a standard by which to measure an emergent best practice).

Comments

• 1. Alan Thomas II: The concept of "wicked problems" may be instructive here: https://en.wikipedia.org/wiki/Wicked_problem (We are attempting a confrontational/collaborative approach, frequently by comparing notes to various "one-shot" solution attempts at different institutions.)
• Shog9: The common concern for recommendation questions is that, if too broadly-asked, they become unmaintainable before they're ever complete (ref). The shopping aspect - obsolescence - is really only a concern for questions where the answer depends on current market conditions.
• 1. Alan Thomas II: As for the second sentence, yes, my point exactly, but a distinction made more here than on the blog or in my experience on Libraries (Main). As for the first, that's an interesting read—and presented as points for consideration rather than a mandate from above, which I will admit improves my mood as the putative subject expert—and I totally agree with most or all of it *in the context of Gaming.SE in general* but much less in the context of the questions that have been asked here. [cont'd]
• 1. Alan Thomas II: They're arguing against "Name everything that meets qualification X" in a field with a **LOT** of potential items, which is undeniably problematic. No argument there. I'm telling you, as a professional expert who's asked these questions here, in this subject, that the question "Which product/service best meets need Y, as defined by requirements A, B, and C" in a field where I can frequently count the potential answers on the fingers of one hand is going to produce a manageable list of results, some of which are objectively better than others, which will be useful to anyone with the same need.
• 1. Alan Thomas II: Unfortunately, neither your nuance nor mine is universally applied here, and a link to a blog post or a Meta question on another site is frequently used in lieu of any discussion of nuance or appropriateness to LIS in any question I or anyone else has brought up of the type "Should we allow question type Foo?" here. The attitudes of some of the people in these past discussions have occasionally been paternalistic and arrogant in ways that have made me more cross than anyone I've been talking to recently (e.g., you or jonsca) deserves, for which I apologize.
• Shog9: The reason I linked to the post on Gaming wasn't to encourage you to adopt *Gaming's* rules - it was to illustrate the importance of discussing problems specific to your community *within your community*. The larger this network gets, the more sites that go through this process, the more we learn and the more lessons we can apply to new sites - but where the rubber hits the road is right here, on *your* meta, in *your* discussions with *your* people. What you decide now will have far-reaching consequences, and we'll do our best to guide you - but ultimately, it's on you.
• Shog9: For what it's worth: Literature.SE took into advisement the experiences of other sites and opted to allow recommendation questions that followed a reasonably strict set of rules. They came to regret this
• the questions garnered a fair bit of participation, but relatively little *interest.* Which brings me back to the question of what sort of questions you want to see *answered* here.
• Ashley Nunn: To support Shog's point - I was really active on Literature before it closed, and the recommendation questions really drowned out everything else that the site had to offer, and quickly became an avalanche of questions that really didn't help the site advance.
• 1. Alan Thomas II: Policy questions, best practices questions, metadata questions, semantic markup questions, the applicability of the *Brandenburg* standard to materials advocating violence, and half a dozen other things I thought of that all fall under "best practices" and are therefore generally subject to the poll / lack-of-canonical-answer / subjectivity-resulting-in-argument exclusions. (I'm trying to think of a way of wording the *Brandenburg* question that would result in an answer that's useful but didn't constitute legal advice, which we cannot legally give.)
• 1. Alan Thomas II: Please note that policy questions and best practices questions are functionally identical for the purposes of this discussion and the next two items on my list are actually both allowed (I'm just the only one who cares :P).

Answer by Ashley Nunn

I am going to start with answering the question that you asked in your title. We can’t change the way Stack Exchange works as an established network of question and answer websites, for a variety of reasons. We are not a forum, and thus there are various post types that aren’t going to work well here, no matter how much we want them to.

Shog already talked about the confusion around shopping rec questions, which is awesome.

I k one of the bigger problems we are looking at here is the idea of Stack Exchange not being a forum. A lot of what people seem to want to know revolves around how different libraries do things, which is fantastic and yay, but is also (generally) more suited to a forum setting than a Q and A setting. Some of the things that make it different from a forum are the idea that we want answers to be suitable to a general audience, not just the very specific needs of one poster, and we are also not really open to extended discussions and people’s opinions on things (unless they have fact to back them). There is a stronger focus on actually solving problems rather than just having open ended discussions on the topic.

I have written about this idea before and maybe revisiting some of the answers there might help make this clearer. Robert’s answa> talks about how

the anecdotal style question you mentioned — one that seeks to poll the community — is where you tend to run into the most problems. What is your favorite [X]? What does your library do about [X]? Even if the the answers are not literally one-liners, those types of questions generally do not make great questions for this type of Q&A.

This is precisely what we are running into. Generally, these sorts of questions are awesome in a forum or aaxed getting to know you type setting, but here, where we want to be able to accept answers and such to questions, it becomes impossible - how can I upvote one library's way of doing things over another? Just because they are different, can I say one is inherently better than the other?

From the blog post Robert mentions - "real questions have answers, not items or ideas or opinions." If your question is more of a collection of ideas or opinions, it's not going to work well in this format.

Stack Exchange and this sort of model isn’t for everyone, and it isn’t always suited to everything people want to ask, because of the nature of how things work. It’s an unfortunate truth, but like most things in life, it is impossible for Stack Exchange to be all things to all people.

Comments

• Gem: I see what you're saying. The big advantage I see with the stack exchange model is the ability to upvote suggestions. That's not something I've seen in most forum formats or listservs. However, if the questions don't fit with the community, I suppose they don't fit. I'm just a little concerned about the viability of LIS stack exchange because I'm not seeing a huge number of questions I would consider "expert-level" questions (at least in the public library arena). Then again, I can't think of any to ask in this format either so I'm not sure how to fix the problem.
• jonsca: Thank you for posting this Ashley!
• 1. Alan Thomas II: I agree with your answer insofar as I agree that SE, as it is configured on most sub-sites, isn't for everyone or every question. Indeed, it is apparently so ill-suited for the questions and methods of much of our profession that we could increase our volume by 400% and still be below the "needs work" marker. This is not a site that would be doing fine if we'd just stop whining or being the wrong sort of person; if the status quo really is the best option on the SE network, the SE network is not the place to have an LIS site. This site is not healthy as-is; what do you suggest we do about it?
• Ashley Nunn: Other than trying to improve the site, get more users who are active and asking and answering questions that work within the model we have, there isn't much else we can do. We can try to teach people how to use the network, but ultimately, if people don't want to use it, they won't. It's not about "the best option" - it's about not changing the basic workings of the network just for one specific site.

Answer by Mary Jo Finch

The question asks whether the LIS Stack Exchange has to follow other sites in terms of appropriate questions. I participate in another forum (Movies & TV - very entertaining), and almost any question on there can be answered by a librarian using his or her fine research skills. I have answered oodles of questions on movies I have never seen, because I can research.

The folly of creating a Q and A site for librarians that seeks canonical answers is that it would be a website for lazy librarians who want others to do their research for them. What librarians NEED is a site where we can get first-hand information from each other about what is going on in other libraries, what works and what doesn't - stuff you can't find in some definitive source. Maybe our questions won't yield a lot of "accepted" answers - because many answers are helpful to the kinds of questions we ask. (We may have to appeal to Stack Exchange for some exemption on the stats in this arena if we are ever going to graduate out of Beta.)

Stack Exchange's mission is to create "libraries of high-quality questions and answers." Librarians are in the unique position of already having libraries. If the information we are looking for has been written about already, we can probably find it without asking each other.

Comments

Answer by gmcgath

I posted a question of my own about "shopping" questions, then saw that this discussion covers much of the same ground. What I'm getting out of the discussion is that only questions which have strictly factual answers are appropriate. Yet doesn't "Is there free software which does X?" have a strictly factual answer? The questions currently up include "Is X a viable option," "Effective presentation," and "Are there benefits," which don't sound as if they lend themselves to purely factual answers, yet they weren't disallowed. I'm really confused at this point.

Comments

• Ashley Nunn: In the case of your particular question, there is no answer to the question in its current state that is going to result in anything other than a list of different software you can use. This is something we want to avoid, as the answers quickly become obsolete as technology changes, and also, it is going to become simply a list of what individual users prefer - there is no real acceptable answer in that case.
• Paul Wheatley: I guess this assertion is not agreed with by many of the LIS users, as was seen with the reaction to the closure of this question (unfortunately much of this reaction was deleted, leaving no record to learn from). I'm concerned that an overly strict interpretation of the "shopping" question rules is not helpful or suitable for LIS. Software focused answers will by nature be a big part of LIS. If they are not possible, this forum is likely to be of limited use to much of the LIS community.