Revise, Rewrite and Proofread.
You should always plan on doing some revision and rewriting. But how much? And how do you know when it is necessary? Here again, if a friend (or enemy) has trouble understanding your paper or any part of it, you need to do something about it.
You need to rewrite the foggy and fuzzy sections. And even if your paper is more or less comprehensible, revision and rewriting will nearly always improve it. Basically, a sense of when and what to revise, what to throw out, and what to rewrite is developed through the practice of writing and through receiving feedback on your papers.
With your research problem, outline and the limitations of readers firmly in mind, go through your outline with a certain ruthlessness. Cut out any jumbles of excess or imprecise words. Don't be afraid to throw away sentences and whole passages which don't do the job of communicating your ideas. Clean up the grammar. Rewrite as necessary.
Now you can type or print your final draft. And then you should proofread it. You don't want to leave any little but distracting errors or typos uncorrected. (A typo can change the comment "Kroeber's theory is not considered adequate" into "Kroeber's theory is now considered adequate." But any type can eb distracting.*) Most word processing programs will check the spelling, and even simple grammar, automatically if you ask, so use these features. However, always read the paper yourself before you print the final copy--the Kroeber typo cited above is neither misspelled nor grammatically incorrect, but it is a mistake that will be noted.
*Distracting typos deliberately inserted.
