Logical fallacies are errors in reasoning. The use of them, whether deliberate or not, can cause a reader to reject your point of view. They are often the result of a lack of evidence or careful thought on a subject. You should check your rough drafts carefully to avoid them. Here are the most common types:

Hasty generalization--the writer bases the argument on insufficient or unrepresentative evidence. (For example, say you wanted to find out how men feel about the new car ads directed toward women, and you interviewed only members of the men's movement. To say that men feel those ads are sexist based only on that evidence would be a hasty generalization.)
Non Sequitur ("it doesn't follow")--the writer's conclusion is not necessarily a logical result of the facts. (Conclude that because Jane Smith is a brilliant historian, she will therefore be a brilliant history teacher is a non sequitur.)
Begging the question--the writer presents as truth what is supposed to be proven by the argument. (For example, if you say that "dangerous pornography should be banned," you are begging the reader to get something for nothing. That is, you are giving no evidence for what must first be argued, not merely asserted, that pornography is dangerous.)
Red herring--the writer introduces an irrelevant point to divert the readers' attention from the main issue. (You are yelling at your roommate because s/he never does the dishes when it's his/her turn. S/he turns around and says to you "yeah, well, what about that $20 you owe me?")
Post hoc, ergo propter hoc (after this, therefore because of this)--the writer mistakes a temporal connection for a causal connection. (An hour before the rape, the man had bought a Playboy; therefore, because he read the magazine, he raped the woman.)
Argument ad hominem ("to the man")--the writer attacks the opponent's character rather than the opponent's argument. ("Dr. Bloom can't be a competent marriage counselor because she's been divorced twice.")
Faulty authority--the writer uses ordinary people, or lay people, as evidence for a specialized area of knowledge. (You interview your roommate, an average college student with no special knowledge or skills, about what he thinks about gays in the military and use that as evidence that they shouldn't be there.)