Let's clarify something: this is a free Q&A site where everyone participating—and mostly those who answer questions—volunteers their time to help others. It's the actual process of doing so that should keep users doing it. Why else would I be spending my time here? And that does not even include the possibility to learn something yourself by researching problems.
From my experience, the amount of users who are not serious about their questions is considerably low. And in fact, those who aren't can be quickly identified as trolls. Trust me.
Considering that, you cannot know if your answers don't get viewed. Unless the OP posts a question, and then never logs in again (granted, this does happen), your answer will probably be seen by others. This doesn't mean they will necessarily vote for it. Maybe they lack expertise to judge it. Or maybe they're just visitors who cannot vote. Keep in mind that 90% or more of our traffic comes from search engines—not logged in users!
And please, do not choose questions to answer based on the reputation of the user. I think everyone equally deserves a chance to get their problems solved, with the only prerequisite being that the question is good and answerable.
If you do, however, see questions that show no research effort, or are unclear (i.e., those that would require a lot of guesswork on your side to come up with a solution), then you should downvote them, and possibly flag them so that they're put on hold. Leave a comment asking for clarification—and only then, if the question is edited into better shape—feel free to post an answer.
Going even further, if you see one particular user who is possibly abusing our system to get answers to their tasks which they're just too lazy to do, then please let us know so we can verify this abusive behavior.
It's your choice how detailed you want to make your answers. If you really go out of your way to provide working examples (the equivalent of JsFiddles on Stack Overflow), that's actually much appreciated. It's not strictly necessary, but in the end it should pay off. Maybe not in the number of votes, but in the possible number of people you'd help.