6

Questions such as Gallery of Step1, Step2, Step3, … progress bar UIs for web apps? are posts requesting lists of examples of a certain type of technique being used 'out in the wild'. I feel these sort of questions are beneficial to the UX.SE community as a whole, as a large part of our actual jobs is competitor research, gathering evidence and sourcing examples of types of technique to use in various situations.

However, these questions do not have a right or wrong answer, they're just collections of links. Should we allow these posts requesting lists of examples (possibly re-enable the Community Wiki option for posters?) so that we can start collating these list-type questions within UX.SE? They could potentially bring people into the site who have been searching for examples of specific UI techniques.

  • I could see some uses for "pattern library" posts though this one is a bit poorly formed; title should be more clearly referring to the type of pattern, and I don't see why "web apps" is relevant for the pattern. – Ben Brocka Feb 23 '12 at 14:41
  • Agreed, it's not a perfect question illustrating my point, but it's the question that triggered the thought. Maybe see if there's a consensus as to whether this sort of question should be allowed, then we can bash out requirements of what should/shouldn't be allowed. – JonW Feb 23 '12 at 14:45
  • You do know that lists are NOT what Community Wiki is for, right? – Charles Boyung Feb 23 '12 at 15:50
  • @CharlesBoyung Yes, but then that's one of the reasons I'm asking the question really. What can we use to display these type of questions. If CWs shouldn't be used for them, and standard Q+A questions don't work well either then what options are out there for these type of questions? – JonW Feb 23 '12 at 15:58
  • 1
    @JonW - Maybe we shouldn't be displaying these types of questions. Not all questions about UX belong on UX.SE. Just like not all questions about software development belong on SO. – Charles Boyung Feb 23 '12 at 16:01
  • 1
    Fair enough opinion, but WHY don't you think they belong here? We want to get as many people to visit the site as possible, they're questions that are in the correct industry, and they're beneficial to us all, so personally I'd rather have them here than have someone set up a seperate site, or post them to Quora or something. Then we'd lose that potential traffic. – JonW Feb 23 '12 at 16:50
  • @JonW - There are LOTS of questions that are relevant to the UX industry that still don't belong here. This is a Q&A site, not a poll site, not a critique site, not a "what <person who shall remain nameless so mods don't delete my comment> thinks this site should be" site. And I disagree completely that "they're beneficial to us all". I find most of these lists of data particularly useless. I can do a simple google search and get almost the exact same results. This really IS why they killed CW (essentially). SE sites are for expert Q&A, not for anything else. – Charles Boyung Feb 24 '12 at 5:08
  • @CharlesBoyung OK, I can see where you're coming from. Can you summarize your thoughts and leave it as an actual answer that people can respond to, as this question comment thread is getting too long now. – JonW Feb 24 '12 at 9:52
8

I'm going to skim the Community Wiki aspect first and get that out of the way - if you decide allow these, don't put Community Wiki in the equation. The issues inherent with list questions are not really about the reputation earned - its major impact is on votes, and Community Wiki doesn't do anything about votes. So if you allow these on the grounds you want to show these off, then people should reap every benefit they get from contributing.


My personal recommendation, in line with how it generally falls across the network, is that you shouldn't allow these questions asking for a list. Recall that your purpose as a site, moreso than anything, is to help people and improve the internet. Your means to accomplish this is by providing solutions to the problems users have.

The solution to a list question is a complete list of all the items that could possibly apply to the author's needs in an expert manner. This is something that has never really been successfully accomplished with how these list questions play out.

This isn't to say that these kinds of questions can't be adapted here. The trick is to change their focus so they can be solved. I'll take the examples from Justin's answer of what questions could fit, and show a simple conversion to change it from "Give me a list of options" to "I need a solution".

  • what are good examples of colored buttons that work well for elderly users?

    Let's instead try something like, "How should I design colored buttons geared towards an elderly audience?".

  • where can I find examples of "sorry, there was a server error" pages that minimize user frustration?

    This is easy to change to "How can I minimize user frustration in server error pages?".

  • what are examples, other than Bing, of putting search results inside search autocomplete popups?

    This one might feel off a bit, but I'd go at least in the direction of "How should I design the output of search results inside search autocomplete popups?".

This kind of transformation may seem like it doesn't change anything, but it changes the things most meaningful to the site while retaining the ultimate goal. We allow multiple answers on the very grounds that, sometimes, there isn't just one single great answer. Sometimes there are multiple solutions. So when someone asks a question where there is more than one possible answer, then it will attract those answers. In these scenarios, I've changed the question from simply seeking examples to more asking about the approach. This would allow your answerers to not only teach the theory behind good UX design to solve those problems, but still allow the full freedom of displaying examples of that theory.

The parallel to how it works on Stack Overflow is simple to understand. If you want to accomplish something, and probably want a plugin or script that accomplishes it, you don't ask for a list of scripts to try and suss out the best option. You ask about the problem you want to solve with the script, and in the end, if the best answer is a script, it will come as such. And, likely, it'll be fleshed out into an awesome result that your readers will like to read. If it gets upvotes, it won't be because it's merely some item that people like, but it'll be because it gets the desired job done.

The answers you get will come in the same flow as you would perceive a list question. But the founding difference is that you can have a conclusion to such a question. Even if more answers may come, a single answer can solve the problem posed by the question, allowing the author to bring closure and allowing future readers to see this as a solved problem. That's what users want to see - they want results. So focus these kinds of things on the results.

Not all lists can be transformed like this into a usable question. But these also tend to be examples of stuff which you would not want to see.

  • 2
    Agreed, I find the specifc focus of the "how" questions results in more useful answers anyway, as it requires people to elaborate on why something works or should be done, instead of just listing a bunch of sites that have pretty buttons/ect – Ben Brocka Feb 28 '12 at 14:20
  • Yes, good points. Having had some time to think on it and from hearing the opinions of people here I think that the closest we should possibly allow to "show me a list of examples of X" questions should just be requests for pattern libraries for specific topics (although even these won't be appropriate all the time) and any other type of questions will need to be amended to a 'How' version, if possible. – JonW Mar 1 '12 at 8:59
  • I giggled when I realized most of those sample questions would've been removed for "Not being constructive" – David Mar 4 '13 at 7:11
1

Yes, "find a pattern library" questions should be allowed, with the caveat that they must meet the same criteria as other UX.SE questions, per the FAQ and the comments above:

  • For user experience researchers and designers.
  • Practical, answerable questions based on actual problems that you face.
  • Eliciting definitive answers or solutions rather than prolonged discussions. For pattern library questions, I think this means requesting links to existing pattern library resources (online, in books, etc.), not asking the UX.SE community to build a pattern library on-the-fly within their answers.
  • Reasonably scoped. For pattern libraries, I would say that "progress trackers for wizards" is reasonably scoped, but "buttons" is too broad.
  • Not chatty or open-ended. For pattern library questions, I think this goes hand-in-hand with reasonable scope. If scope is small, then there will often be few good pattern libraries, making a "best answer" less subjective. For example, AFAIK there's only one good pattern library for wizard progress trackers, so choosing a best answer would be easy.
  • Provide context: mockups, platform info, constraints on implementation or design, etc.
  • Not asking for icon help

To iilustrate, here are some examples of "good" pattern library questions:

  • what are good examples of colored buttons that work well for elderly users?
  • where can I find examples of "sorry, there was a server error" pages that minimize user frustration?
  • what are examples, other than Bing, of putting search results inside search autocomplete popups?

And bad examples (close these questions!):

  • where can I find 100+ examples of pretty orange buttons?
  • is there a site where I can download free wordpress themes?
  • what's a good place to get CSS code samples for different kinds of popup menus?
  • My colleague hates gradients. Where's a list of sites I can use to prove him wrong?

I don't think Pattern Library questions should be Community Wiki or any other special kind of question as long as they meet the criteria above. This is just like how "find a UX book" questions are handled today on UX.SE. Overly-broad or non-professional "find a book" questions like this one are closed or community wiki-ed, while more focused, practical, and narrower-scoped questions like this one are allowed.

I also don't think that pattern libraries need to be useful "to us all" to be allowed, any more than any other question needs to be useful to everyone. It just needs to be useful to enough UX experts that it's worth taking up space.

I also don't think that ease of finding in Google should disqualify a question. We should give askers the benefit of the doubt that they've tried for a reasonable amount of time to find the answer on their own, and if one person can't find the answer then it's likely that others have the same problem.

Also, many things are hard to find in Google for odd reasons. For example, if you search for "wizard progress tracker" in Google, you won't find the wizard progress tracker pattern library because the word "wizard" is never mentioned in that article, just like the word "wizard" never shows up in the UX.SE question about wizard progress trackers. We shouldn't penalize UX askers who happen only know the "wrong" synonym from Google's point of view.

Full disclosure: I wrote the wizard progress tracker library question that inspired this meta question. But my answer here is intended to be a general defense of focused, appropriately-scoped requests for references, regardless of whether or not my quesiton met those criteria. :-)

  • Community wiki answers don't bring any reputation point to the poster. That's why pattern library questions should be marked as such. – dnbrv Feb 28 '12 at 23:36

You must log in to answer this question.

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