Is sharing prior research does more harm than good, in general, in Q&A sites?
Some Q&A websites such as StackExchange require from people asking questions to share "minimal prior research" that they had done to try to answer their question themselves (and often also juxtapose a request for "evidence" for such prior research).
I can understand why "prior research" is required for say, Code review StackExchange, where I myself shared several codes I have developed after I first tested them myself and understood that they allegedly work as expected but I still wanted a review to make others and myself learn if perhaps I could write the code more efficiently.
- Prior research enables community learning
- Prior research is good for SEO as it raises the quality of a webpage; simply put, the webpage becomes more "on topic" not only by a (desirably) good heading but also by a richer inner content.
But if:
X = My Problem
Y = How I tried to solve it but failed
Z = My question (why how I tried to solve it failed?)
One could easily be accused with the (philosophical idea) commonly known as "XY problem".
And yet, if we only describe a problem, perhaps the only practical question to ask without causing an "XY problem" would be "How would you solve that problem?" and if so, perhaps a better concept than "Q&A" websites would be "Problem describing and solution suggestion websites".
Is sharing prior research does more harm than good, in general, in Q&A sites?
2 answers
You are accessing this answer with a direct link, so it's being shown above all other answers regardless of its score. You can return to the normal view.
I think that prior research might not be bad if the OP:
- Describes a problem (in understanding or with material) without asking a question in heading (
H1) - Asks "how would you solve that problem" in the end of the post
So I think that generally, questions should not appear in titles at all (although it is very tempting that they would from an SEO standpoint) and that between number 1 and number 2 should come a description of what was done so far (if at all).
0 comment threads
I think the point of requiring evidence of prior research is to avoid clogging up the site with endless repetitions of the same basic questions. The point of a Q/A site, or, at least, the stated goal of Stack Exchange, was not to provide answers to individuals to to build up a permanent resource that would be of value to many people asking the same question (a much more commercially valuable resource than a random web forum). To that end, they preferred to have each question asked once, not a thousand times, and thus they wanted people to look for an existing answer before asking a new question.
The problem with this is twofold. First, that is more work than just asking you question without doing any research, and people generally take the path of least resistance. Second, the same question can take many different forms and people may not recognize their question when they see it asked a different way.
And, it turns out that, because the same question can be asked in different ways, there is an advantage in having it asked in all those different ways, even if most of those ways don't attract the best answers, because at least people can recognize the question they want to ask. But this creates many pages with individually low SEO as opposed to one page with high SEO, which is less commercially desirable.
There is also the related problem that two or more completely different questions can have the same answer. One sees this all the time on Stack Overflow, where question are marked as duplicate because they have the same answer, even though the question are not remotely alike and people with one of those questions would never recognize their inquiry in the second question, and would therefore never associate the given answer with their question.
The broader research problem, though, is that formulating a clear question is often more than half the battle, and once the question becomes clear, it is often quite straightforward to find the answer. As often as not, the real problem people are struggling with when they post on QA sites is formulating the question correctly, or diagnosing the problem correctly, rather than finding the solution. Thus again we get duplication, and the same basic question asked over and over again, because people don't know what question they are really asking.
This seems to me to be particularly true of writing sites, where most writing problems boil down to a handful of basic principles of grammar, style, or story structure. What the best answers do on this site, over and over again, is to say, your question is an instance of this type of issue, which is addressed using this general principle.
In short, doing that basic prior research is often the hardest part of the problem, and the thing that the questioner is having the most difficulty with. So, the requirement for research, though it makes perfect sense in terms of trying to build a permanent resource of the best questions and answers, fails to recognize the fundamental difficulty of formulating the right question in the first place.
Thus, rather than building a permanent resource of the best answers to canonical questions, QA sites end up as a question categorization service for individuals, which is a valuable service, if one that is perhaps not as commercially desirable as the original goal.
0 comment threads
1 comment thread