1
$\begingroup$

This question has been marked as a duplicate of: Can we encourage people to put more research into questions?

While the linked question is about making people actually think and read up on the topic they ask about, this discussion is about finding ways to make people use site-tools and check out the automagically proposed questions when they write up their own questions!


Recently whenever I browse questions I see at least one if not more that could be answered or at least improved by going through answers to previous questions we've had on the site.

E.g. this question about mammals with blue blood. Reading the question I immediately had two thoughts:

a) Have you actually googled this? Tried to read up on the topic?

b) You have found this stack, have you tried looking if the question has already been asked?

I know at least the answer to the latter must be no, because I know this question has come up already, and I know this question has been answered amazingly well here (and not only about blue blood).

What I also know is that searching is hard. Terms & words that I use when thinking about a question might not necessarily match those used by others. When searching I have to look up existing tags, and synonyms of words I would use - I have to be ready to do that extra effort. But I know that the effort is worth it, because it helps me with whatever I do or try to do eventually.


Searching the meta for the topic I find exactly two questions/discussion on the subject over the last 4 years:

A meta question about what to do with questions that show obvious lack of research

A meta discussion about how to properly search for existing questions


How can we incentivise people to use tools like the site-search or related questions before and while asking questions?

$\endgroup$
  • $\begingroup$ Possible duplicate of Can we encourage people to put more research into questions? $\endgroup$ – Renan Nov 17 '18 at 10:10
  • $\begingroup$ @Renan addressed $\endgroup$ – dot_Sp0T Nov 17 '18 at 10:29
  • 2
    $\begingroup$ Related opinion: My line for "not researched" is quite simple. If I take the title text of your question, verbatim, and paste it into google and one of the top three links explains the answer clearly (typically wikipedia), then you haven't researched your topic. Unfortunately for my approach, voting to close with a link from Let Me Google That For You is considered rude. $\endgroup$ – Cort Ammon Nov 17 '18 at 14:25
  • $\begingroup$ @CortAmmon that would be the linked question as well. $\endgroup$ – dot_Sp0T Nov 17 '18 at 14:30
  • $\begingroup$ @dot_Sp0T If I'm reading it right, the linked question just incldues the first part (put it into google). The part that has gotten me in trouble in the past is the use of LMGTFY $\endgroup$ – Cort Ammon Nov 17 '18 at 14:32
  • $\begingroup$ @CortAmmon I mean more that I was hoping for this discussion to go into the direction of finding out why the site-search and functions like 'related questions' aren't put to more use by most querents. And even more hopefully finding ways to put these functionalities into the minds of people.. $\endgroup$ – dot_Sp0T Nov 17 '18 at 14:37
  • $\begingroup$ @dot_Sp0T Do you find this to be an issue with just new users or users in general? It's harder to encourage this with new users because they don't understand the process that underlies Stack Exchange. And obviously they need to decide to do this search before asking the first question. $\endgroup$ – Cort Ammon Nov 17 '18 at 14:47
  • $\begingroup$ @CortAmmon in both actually. New users (any actually) get propositions on similar questions while writing them (based on the words in their title and tags). When answering a question you first have to read it, which means next to the question text is a list of related questions as determined by the algorithm based on words, tags, etc from the question, comments, answers. All are generally ignored.. $\endgroup$ – dot_Sp0T Nov 17 '18 at 14:51
  • 1
    $\begingroup$ I am delighted by the irony inherent in voting to close this question as a duplicate. $\endgroup$ – Dewi Morgan Nov 17 '18 at 17:05
  • $\begingroup$ I believe this page is shown for new users when asking their first question? If they don't want to follow that, then I'm not sure they would follow any other advice... $\endgroup$ – Andrew T. Nov 17 '18 at 17:52
  • $\begingroup$ @CortAmmon -- Frankly, I wish that were one of the drop-down responses in the VTC list! I don't think it's any more rude to point out that a querent has obviously not even googled his own question than it is to spam this forum with an ungoogled query in the first place! I've only ever asked one question here, and believe me: I searched as many different key words as I could think of both externally and within WB.SE itself! $\endgroup$ – elemtilas Nov 19 '18 at 3:04
4
$\begingroup$

Personally, I set the bar low. Often the issue is that the asker doesn't know what to ask for and is struggling. Other times the asker has some tiny implicit assumption that they have no idea how troublesome it is, and needs help dislodging it. I find this happens quite often in virtual reality, teleportaition, and time travel questions, where the reader implicitly assumes their intuitive understanding of what an "individual" or a "self" is will hold. This intuition ties them in knots which can't be untangled, but a slightly more nuanced definition of a "self" often causes the Gordian knot to unravel completely. (As an aside: one of the theories about why the Gordian knot could not be untied is that it was actually an unknot. This would explain why the most brilliant of thinkers couldn't find a way to untie the knot. It was already untied! As legend has it, Alexander the Great finally "untied it," by taking a sword out and cutting it in two!)

My bar is simple: If I take your title text, and copy it into the WB search, or copy it into google, and one of the first few hits is obviously a clear match with a great answer, you haven't researched it.

In this particular case, I actually think the question passes the test. Other Blood Colors was more focused on the chemistry. It explicitly stated to ignore the biological problems that a change of blood color might have, and focused primarily on the chemistry of blood oxygen transport. Would it be possible for mammals to evolve blue blood is more of a practical question.

The winning answer in Other Blood Colors does mention that there is a 50% efficiency drop when you switch from hemoglobin to hemocyanin, but goes no further. I think a good answer for "... evolve blue blood" might reference this, and go into why a 50% drop in efficiency would be a big deal for warm blooded vertebrates. Cold blooded might be an interesting question to explore.

As for how to actually make things better, I think that's just culture. Mark things as duplicates, and possible duplicates. I often will put a possible duplicate in a comment rather than a VTC, just to see if that was indeed the same question, or if the author just hadn't yet phrased the question in a way which was distinct.

$\endgroup$
2
$\begingroup$

The trouble is that the SE/SO search system isn't particularly intuitive, and certainly isn't as intelligent as a google search. When you search for something, that's what you get. If you don't use the same words as someone else, it doesn't do a fuzzy search so you won't get the response you're hoping for without some intensive hunting.

What this means in the long run is that we want to build up a database of duplicate questions. If those duplicates are correctly linked through to our definitive question on the subject then the vague searches will, in time, lead people through to the correct question.

Those of us who have been around longer are better at knowing that something has been asked before, but mostly because we saw it in person the last time, so we know it's definitely there to be found with the right search terms.

Also, are you seriously asking how to make people think before they speak? You know that never works.

$\endgroup$
  • $\begingroup$ But at least make them look at the proposals. In my experience the SE suggestion engine bringing up possible duplicates when VTCing, as well as proposing questions about the same topic when writing your Q, as well as proposing related questions on the right side once it's posted ALL work very well $\endgroup$ – dot_Sp0T Nov 23 '18 at 12:14
  • $\begingroup$ @dot_Sp0T, the duplicates in the VTC system have been manually added. The possible duplicates while asking is a much better search than the actual search, but it's still very dependent on how you (ab)use the language. $\endgroup$ – Separatrix Nov 23 '18 at 12:19
  • $\begingroup$ So you tell me all the dupes in the list showing up when I am the first to VTC on a brand-new question have been manually added before I press that button..? :/ $\endgroup$ – dot_Sp0T Nov 23 '18 at 13:00
  • $\begingroup$ @dot_Sp0T, I don't get a list unless they've been added. It could have come from a flag rather than a VTC. $\endgroup$ – Separatrix Nov 23 '18 at 13:51
0
$\begingroup$

Ignoring the specific questions...

There is no way to enforce this action

No mechanism exists to encourage better research. No mechanism exists to guarantee or ensure better research. We're stuck with the fact that anyone who wants to can post anything they want: properly researched or not.

We literally have only two options:

  • Downvote (the appropriate action)
  • Close (theoretically only necessary in extreme or clearly defined circumstances)

Personally, I rarely downvote and try to avoid closing unless I'm sure leaving a question open is going to lead to a bigger mess. My personal preference is education. It's a neverending duty because newcomers are always joining — and I cannot trust that they've even bothered to read our rules (indeed, quite the contrary, I can generally trust that they haven't read a single thing about our site other than its name).

However, having said this...

It's important to realize that what's 100% obvious to you isn't 100% obvious to me or (necessarily) anyone else.

  • We have a lot of English-as-second-language people using this site, and they tend to take a beating simply because their English is poor. That's really unfortunate, because their imagination isn't. What is easily researched in English may not be easily researched in (e.g.) Swahili at all.

  • We have a lot of youngsters using this site. I do not mean that in a derogatory way. I'm in my 50s. I have a degree in engineering and a lot of life experience. A considerable number of users don't have a college degree and don't have much (if any) life experience. Expecting them to understand what I understand is arrogant at best, conceited at worst.

And speaking of arrogance...

  • We have a LOT of people on this site with amazing abilities, skills, and education. There are, for example, people on this site who know more about orbital mechanics than I ever will. Why is it important to know this? Because if I don't know my limitations, then I will be a terrible judge of other's limitations.

Therefore, while I understand your frustration completely, I must advocate patience. Patience that would make a successful parent of septuplets appear rash and unforgiving.

$\endgroup$
  • $\begingroup$ Again, this is mean as a discussion about the site-functionalities that actively propose other questions while you, e.g. write the question. as well as afterwards once you've written it. there's a plethora of proposals and they get almost always ignored $\endgroup$ – dot_Sp0T Nov 23 '18 at 6:21
  • $\begingroup$ On other sites I've proposed changes to what is presented on-screen during the question-writing process to help guide question writers. The fundamental problem is that nearly everything on the submit-your-question screen is controlled by SE, not the individual sites, and therefore is a massive pain-in-the-tuckus to get changed. If a systemic improvement is what you're looking for (I didn't read your question that way, sorry), you need to ask this on Meta.SE, not here. $\endgroup$ – JBH Nov 23 '18 at 14:57
-2
$\begingroup$

The suggested answers are often totally unrelated (at least by title). They have similar words and phrases in them, but don't have anything else in common.

So the first step to making people actually look at the suggested related questions is to actually present related questions.

What I found interesting when I was very active on another SE site is that I often knew that a duplicate question existed, because I had already answered that very same question, but was unable to find it through the site search! Which tells you a lot about how powerless automated searches are.


If I knew how to search for what I wanted to know, I wouldn't be here. Often I don't know what the thing I want to know is called. And I never know the solution, which is usually what the subject is that is found in searches.

For example, if I ask about reasons for aliens to attack Earth, I might be asking about psychological or social issues. Wars are begun to pacify political unrest, so I might be asking about the psychology and social aspects of politics, without being aware of it. My answer might be found by searching for "psychology of war", not for "why do aliens attack Earth".

If you know the answer, these relations seem obvious to you, but to anyone who is looking for an answer, it is often not clear what good search terms would be. Which is why using the site search (or Google) often turns up nothing relevant at all – or the actual answer hidden on page 63 of millions of search "results".


My answer would be:

Show related questions that are actually related.

As long as you cannot, it is totally fine not to research at all.

If you know that the same question has been asked before, mark it as duplicate and link to the first question. That will take you much less time than it would the asker to find that duplicate.

If you don't know that a duplicate exists, and you want to answer the question, then quite obviously repeating the question has brought it to your attention and given you a chance to answer either the duplicate or, if someone else marks it as duplicate, the original. Therefore asking the same question again was a good thing!


Finally, if it tires you to answer the same question again and again, then maybe you need to work on yourself, or you need to move on. Think of a teacher who told his class that he answered the same question yesterday, or that his pupils could find the answer by using Google. Not only would he be out of a job if learning acutally worked that way, he also completely disregards the interpersonal aspect of learning. People want to ask a person and have things explained to them. Not (necessarily) because they are lazy, but because humans are social animals. If I have to find an answer alone, then finding that answer is not as much fun as interacting with another person to find it. Which is why I am totally opposed to the concept of "duplicate questions". If I want to help others find answers, then repeatedly answering the same question is a part of that.

$\endgroup$

You must log in to answer this question.

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