There have been a number of questions tagged "physical" in recent weeks. As interesting as the questions are about physical objects and how people use/abuse them, I couldn't help but wish for an answer from someone with an industrial design (or related engineering) background that could cite relevant standards/research/best-practices.

Perhaps it is just my perception of the answers, but these questions tend to attract more than usual speculation about "why" something was designed the way it was, or how to fix it. There are some really good answers, don't get me wrong, but I don't have an industrial design background, so what looks like a "really good answer" to me, may be an obviously unsuitable answer to an industrial designer because of some manufacturing constraint, or historical reason.

I would please like some advice on how to get more exposure to industrial design research, so that I can contribute better to answering that class of questions. I already own Don Norman's "Design of Everyday Things", and I know he has a follow-up book as well. How would one build knowledge in this domain if it doesn't overlap your full-time job?

The flip-side of this brings me to the second part of my original question, where I would appreciate some advice on answering what I term "rabbit-hole questions". The recent question about tag-use by non-technical and non-web-savvy people is a case in point. This is a really hard question. Due to my background and training I am very comfortable in surveying academic research, and I have spent a couple of days collecting and working my way through a multitude of research papers, and two book chapters, covering fields such as psychology, cognitive linguistics, computer science, library science, information architecture etc.

I am still busy trying to reduce all the available research to an acceptable answer, but the short answer is pointing towards "well, it's difficult, first you have to understand how people learn to categorise things, and then consider how tag systems fit into the whole design of your site, and after that it depends on a lot of things that researchers don't really like reporting". I know that I should refrain from blog-post length answers, but how would you provide enough background information to contextualise your answer in such a limited space, and without referencing all 20+ papers that helped fill in the blanks?

3 Answers 3


A couple of points here, I think.

Firstly, you don't have to be able to answer every question that comes up. Nor should you be expected to understand all the answers. UX is a very wide field, taking in all sorts of areas - usability, testing, accessibility, visual design, IA, prototyping... you aren't expected to know everything about all subjects.

Just as on http://stackoverflow.com it's not expected that everyone using that site knows about everything that gets asked. There will be PHP experts who focus purely on those questions, often even using the tags to Ignore various other tags they do not care about or have no knowledge of. That is fine too. In fact that's one of the intended uses of tags. Don't feel guilty that you can't contribute to everything. People who just answer every question going regardless of their knowledge in the area will not do themselves any favours in the long term.

As for your other point - basically feel free to leave as much information as you feel is relevant and that answers the question. You can provide summary information from the primary sources, citing those sources against them and close out your answer with a list of 'other useful information sources'. Provided your answer actually provides an answer itself and isn't just a list of links then it can be up to the reader to decide if they want to dive down the rabbit-hole themselves to find out even more, you are just helping them on their way.

This site isn't like skeptics.stackexchange where multiple sources and detailed answers are a requirement for answers, we just need answers to answer the question as best as they can, and provide additional reading materials for those who want to really get to know the subject. Answers with research are usually significantly better that subjective ones (and the upvotes they receive show that the community values these answers too)

However, bear in mind that this is a Q&A website, not a fully-fledged research tool. It's for people to find answers to questions. A 2,000 word essay is unlikely to be an answer, but would be supporting information for the actual answer. UX practitioners of all people should know that people don't actually read much of the content on websites! However if people want to read more then give them the option to do so.

  • I wholeheartedly agree with the "don't have to be able to answer every question" sentiment. Over the years I have found Socrates to be right when he said "The more I learn, the more I learn how little I know", and the multidisciplinary nature of UX reinforces that statement for me. My personal approach is to try and provide some research-based answers, and the community reception has been very kind. I have come to see the site as an opportunity to be introduced to new problems or re-evaluate my own opinions. It has been a humbling and valuable learning experience so far. Thanks for the advice!
    – CJF
    Feb 19, 2013 at 12:26

I know that I should refrain from blog-post length answers, but how would you provide enough background information to contextualise your answer in such a limited space, and without referencing all 20+ papers that helped fill in the blanks?

Refrain from blog-post length answers where they're unnecessary. If there really is that much research on a topic and you're aware of it, please include it in an answer. Such answers are rare more because they're a lot of work and can be hard to write in a readable way, not because we discourage long answers (we only discourage long answers if they're rambling/unnecessary).

Take this answer of mine for example. With a combination of formatting to draw attention to the important bits (to "hook" people/for those with short attention spans) and links to research I think it's a very good example of a well researched but still quite readable answer.

As for the rest of your points, unfortunately industrial design is more at the edge of this site's current core; an interesting one, but one with few (if any) regular experts. It's also one of those topics everyone is familiar with, even if they don't "know" all that much about the internal bits, everyone has used a calculator or wi-fi switch or whatever.

All stack exchange sites suffer a bit from questions that are very popular with novice users or users from other sites with little specific expertise. They get vastly more attention than they get elaboration, so we end up with lots of "yeah, that sounds right" moreso than "this is what the research says". And, in some cases, I'm not convinced there is always (public) research on the matter. Particularly when discussing why X was done this way, sometimes the answer is just "some guy drafted it that way and they found no reason not to build it".

  • 1
    I think the questions that are "Why X was done this way" are poor questions for this site because unless there is a definitive article from the team who built it as to why they did it that way they're just going to solicit guess answers. Really they should be reworded to "How can I improve on the design of X in a manner that makes it more suitable for Y"
    – JonW Mod
    Feb 19, 2013 at 17:23
  • @JonW Agreed. Those are generally interesting but unanswerable questions, more suited for Quora's model of "well I think.."
    – Ben Brocka
    Feb 19, 2013 at 17:38

As a community, we need to attract and retain new members who have experience with industrial design and can answer those questions. Right now, we mostly have web designers and a bunch of programmers who occasionally dabble with design. This is a problem for our community and has been for a while, but it's not something you can really solve overnight. It just has to happen over time. One effective way to go about doing it is to keep asking industrial design questions so that our site surfaces more in Google for such queries, and questions hopefully eventually get picked up in industrial design circles.

Your third paragraph is a good question that you should go ahead and ask on UX.

Rabbit hole questions (love the term) are another challenge of our field. Since UX is so broad, there are many topics that most of us only have a cursory knowledge of or passing interest in. That's why you see many answers that cite no references or evidence to back up their assertions on rabbit hole questions. In the case of your example question, I did try and spend 20 minutes looking for actual research to post as an answer because unresearched answers are a pet peeve of mine and I don't think they improve the quality of our community. Having said that, I am also no expert on how people understand tagging, so at best my answer is just the result of quick Googling and a nudge in the right direction.

Overall your question touches on some deep challenges for UX.SE that have been present for several years. It would be nice if we'd solved them by now, but I guess they're inherent to being a Q&A site about "UX" and we'll probably continue to deal with them as time goes on.

  • Thank you for highlighting the history of the challenges of UX.SE, as I only joined the community recently, and I'm starting to understand how things "work" as time goes by. Just to clarify, I wasn't trying to pick on your answer :) I was familiar with the papers you cited, but had a look at them again, and it struck me that most researchers only report on how people tag, and not why people choose NOT to tag. The bounty made me look at the work again, and after reading some of the other unsubstantiated answers, I tumbled down the rabbit hole.
    – CJF
    Feb 19, 2013 at 12:10
  • 1
    I know you weren't trying to pick on my answer, I just felt like explaining how I agree with you and was trying to contribute in a similar way. :)
    – Rahul
    Feb 19, 2013 at 13:22

You must log in to answer this question.

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