If i was writing a survey to gauge how many women had been sexually assaulted, and my funding was based on this, it is in my best interest to inflate this number.

So what I would do is ask a question like "has you or ANYONE YOU KNOW been groped or touched inappropriately in a sexual manner?  (This allows for a much wider range of things to be included as I get to determine it all as sexual assault as opposed to what SHE 'thinks'  was sexual assault.  Afterall I wouldn't want her making up her own mind on whether or not she's a victim do I?  This is MY survey).

What this also does is allow one incident to be included multiple times.

Example:

10 girls are on a night out and Rebecca gets groped and tells the dude to fuck off.  She doesn't consider it a sexual assault and she laughs it off with her girlfriends and moves on with her night.

Now the next day in class a sociology graduate gives those 10 girls the survey and asks the questions:

'Have you, or ANYONE YOU KNOW, ever been groped or touched inappropriately in a sexual manner?  Made unwanted physical contact?  Etc.  Etc.

Now 9 of these women saw Rebecca get groped, so seeing the 'anyone you know' part of the question, they all answer yes, including Rebecca.

When you tally it up, you have Rebecca getting some ass-grabbing and hey presto, it's now 10 cases of sexual assault.  Now take that "100%" rate of sexual assault and apply for emergency funding and blast it across campus newspapers/tv media.

Welcome to the American College zoo. This is how easy it is to make up the '1 in 5' bullshit.