Sactoking,
The results do come from % of respondents who partake in the research and answer in a certain way. You are right that it is very difficult to create an accurate sample population, but there are numerous statistical analysis you can perform to verify that your sample is representative of the population you are studying at large. Such a practice is fairly typical, because due to any number of reasons it is impossible to survey everyone. We tend to look at gender, age, education level, race, and whatever other relevant participant factors are needed to determine if the sample differs significantly from the population at large. In the study I participated in, we looked at college students.
Now, I agree, if one study comes up with a 25-33% number (note this includes *attempted* rape as well, not just completed rape) that would seem very high. I was extremely skeptical of this percentage myself. However, the number is fairly consistent. I read probably 30 different research articles, all from different authors, with different methodology, different populations (college, military, civilian, minorities, ect), and the percentages are fairly consistent. We're also dealing with a lifetime prevalence rate, obviously the older you are the more likely you are to experience this situation, though college age women seem to be the most at risk (well, behind the military). Since almost every study finds about 1/4 of the women answer in the affirmative, it seems pretty scientifically conclusive at this point.
In general, if your statistical analysis reveal a less than 5% chance that your results are due to random chance (i.e., what you found was just due to random luck), then we consider it statistically significant.
Researcher's don't just try to replicate others findings. Some do, because repeatable results are important, but it's also important to be skeptical of research findings. Whenever you design a study you are asking why they found something, and trying to change your study to in a way that might explain if one of those alternate variables you mentioned affected in. My study looked at the attractiveness of the victim and perpetrator to see if that influenced third party beliefs about whether raped occurred (ex: if you have a highly attractive perpetrator and a non-attractive victim, is the rape seen as more acceptable?) We get attractiveness data by basically conducting a whole separate study with different participants.
daishi5,
There are definitely a lot of problems with "no means no," however I was using it as a general phrase

As I understand the law, if one person is drunk and the other is not, consent cannot be obtained. One of the more interesting scenarios you get into is what if both of the individuals are mentality handicapped in some way (i.e., lower than normal IQs). Can they consent to have sex with each other? For the situation you described where both people are drunk, I do not know how the law would apply. For one, there are varying levels of drunkenness. I've had sex with my wife when she was "tipsy" and I was sober, I don't think that makes me a rapist. There are just to many additional circumstances that may be involved.
Understand when we conduct this research we do not ask questions like "Were you raped," Google the Sexual Experiences Survey and you will get an idea of how the research is conducted. We tend to look at what we term "sexual victimization," here's an excerpt from a paper I wrote:
"Sexual victimization includes unwanted sexual contact (i.e. fondling), attempted rape (i.e. non-penetrating), and unwanted sex (i.e. oral, vaginal, or anal intercourse). Studies conducted on sexual victimization often find that 30-40% of all female college students have experienced at least one episode during their lifetime (Himelein, 1995; Tetsa, VanZile-Tamsen, & Livingston, 2004). "
There is a continuum, obviously unwanted contact is less serious than attempted or completed rape, yet there seems to be quite a bit of evidence that the a large amount of woman have a pretty bad experience with their bodily integrity at some point.
I do agree that there is a stereotype against male victims. They are not taken seriously, because in our society is not "normal" for a man to refuse sex. Male rape victims have it particularly hard, in part because, for lack of a better way to phrase this, having an erection isn't always directly under your direct control.
I also sympathize with men who are victims of false accusations, because not only does it make the male falsely accused look bad, it weakens the position of those who actually are victims.