The Stack Exchange network suggests, as one of the criteria to determine the health of a beta site, that you should have answers accepted with a good frequency.
I worry that this may not happen much for philosophy-related questions, as most people take the answers as a step in the direction that they are exploring, rather than the final answer.
Should we encourage people to mark answers as correct even if it only helped along the way? If so, how can we do this?
Accepting an answer is not necessarily intended to mark [the] one that is correct, but rather the one that the asker found most helpful.
The official guidelines on accepting answers provide the following "golden rule":
The bottom line is to accept the answer that you found to be the most helpful to you, personally.
Obviously validity is a useful and important criterion for determining the "most helpful" answer. But it is certainly not the only one, especially not on a site such as this one, where the answers are likely to be less objective and clear-cut than those on (for example) Stack Overflow.
But it's really best not to worry too much about the accept rate. The guidelines for a healthy site don't mention acceptance rate at all, and for good reason. By far, the most important thing is voting. That is how we get the best answers (both the most technically accurate as well as the most useful) to the top, and it's how we regard those who provided such answers.
Voting is also important because it's the way the entire community can be involved in selecting the best answers. By contrast, accepting an answer is a privilege granted exclusively to the asker of the question.
So as another question here on Meta implores: Vote early, vote often
I totally agree with Cody. There's another aspect of accepting answers to consider as well. To adapt a post of mine from the Musicians meta:
Accepting answers is completely optional.
The question owner is not required to accept an answer to their question. We view accepting an answer as a simple social convention, a little informal “thank you” between the asker and answerer, a virtual tip o’ the hat to that person whose response, as the question owner, you personally found the most helpful.
That doesn’t mean the community will agree with your choice. But as the question owner, it is your choice to make.
The default sort order is “votes” for a reason. Normally, the best answer will automatically float to the top through community voting. This is important because we expect a lot of our question askers to be drive-bys, programmers who ask a single question, get the answer they need (or don’t), and are never seen again. This is intentional and by design.
From http://blog.stackoverflow.com/2008/11/why-cant-i-accept-my-own-answer
Basically, voting is extremely important. Accepting answer is great, but we shouldn't be pushing users to do it if they're not ready.
Accepting answers early can potentially discourage other answerers.
Keep in mind we have very few active users. We only went into private beta yesterday! As we gain more users, we will gain more and better answers, and the number of accepted answers will go up accordingly.
After much thought, I'd like to propose a set of criteria for upvoting and downvoting answers.
Good answers in philosophy follow the 3 Rs in this order:
Well Reasoned.
Well Researched.
Well wRitten.
Answers (and most questions) that satisfy those properties will earn my vote. Answers that satisfy none of those properties deserve a downvote. The first point is all-important and an answer that is well reasoned has a good chance of earning a vote even if the other points aren't quite covered.
Note that my personal agreement or disagreement with an answer has little to do with my vote. (It's human nature that we will be biased toward answers that agree with our believes, so an analysis of my voting pattern will likely show that I find some types of answers better reasoned than others. We should do our best to avoid bias.)
On the other hand, when it comes to selecting an answer to be the accepted answer, I will pick either the answer that is most reasonable to me or the answer that is most interesting. I'm not much worried about bias in that case.
