Pete L. Clark posted as an answer to a different question:
Here are some tips for improving your posts:
1) Make your title as descriptive as possible. In many cases one can actually phrase the title as the question, at least in such a way so as to be comprehensible to an expert reader. For instance, a recent question on the site is titled Why is the Hilbert cube homogeneous? This is an ideal title: if you know what the terms mean, you probably understand the question already. Among Chandru1's previous titles, Dimension of Vector Space is quite vague (although it is admittedly not so easy to come up with a good title for this question), whereas Cancellation of Direct Products is pretty good, but could be even better: what's the question?
Since you can, and should, put a lot of content into the title, what goes in the body of the question itself?
2) First, in many cases it is appropriate to restate the question in less succinct language, defining or linking to terms as you feel appropriate. (On MO, we like this but realize that it is not required and not even always helpful: if you are asking a sufficiently technical question, it may be best simply to write for an audience which has a certain technical background, which is part of what the tagging system is for.)
3) If the question is homework, you should absolutely say so up front. If the question is not homework, you should take steps to convince us of that. A good way of doing this is by providing context: e.g., I was thinking about Theorem A, and I wondered what happens when you change/weaken the hypothesis / I have done work on Topic B, and this raises the question / I was reading Document C [tell us which one, specifically! link to it, if possible] and I started to wonder...
Especially, if you did not come up with your question yourself but it is taken from some particular source, I would feel most comfortable if you cited the source. One feature of mathematics is that a good question can be a contribution just as valuable, or even more so, as an answer or a proof. You should not try to pass off others' questions as your own.
4) After you give background and state the question clearly, it's best if you say a little bit about what you've done to try to answer the question. If the question is asking for a reference to something, tell us where you've already looked. (I hope you have already tried wikipedia and google). If the question is asking whether something is true, what do you think and why? If the question is asking for a proof of something, what you have tried so far?