I recently had a discussion where I found out that a lot of my friends are terrified of being falsely accused of sexual assault. They think it could happen at any moment and that there is nothing they can do about it.

As if reading from a script, everyone had something to add:

“You know the girl carrying around the mattress? I was reading about it, and it seems like she made it up.”

“I had a friend who it happened to and it’s kind of ruined his life. He was asked to take a semester off without a hearing, and when he came back he found out she had dropped the case. The worst part is that even though I have no reason to, part of me still doubts him just because she said it.”

“It’s like the worst form of slander—your friends stop liking you; you can’t get a job.”

I was shocked. I had never worried about falsely accused men. When sexual assault came up, my first thought was always for the survivor. It has always felt like my duty as a feminist to trust any woman who accuses a man of rape.

In order to debunk my friends’ claims, I began researching false reports of sexual assault on college campuses. I had heard that only two percent of accusations were false, and I wanted to confront my friends with the facts.

I found numerous studies that estimated what percentage of reports are false. Though most put the percentage between 2 and 11 percent, there was almost no consistency.  David Lisak, a prominent psychologist who studies the behavior of rapists, found that six percent of reports were false and 14 percent did not include enough evidence to make a determination of their accuracy.

That study was done at one school: American University. I am skeptical that a single school can represent every college in the country, only included cases that were reported to the university’s police.

I was also surprised to find out that there is very little consistency in how consent is defined; American University calls sexual contact non-consensual if one party is “under the influence of a controlled or intoxicating substance.” Any amount of alcohol or drug, then, could arguably prevent someone from giving consent.

Bowdoin’s policy says that people cannot give consent in a state of “incapacitation” that prevents them making “informed, rational judgments. States of incapacitation include, without limitation, sleep, blackouts, and flashbacks.” With different samples of women and different versions of consent, it didn’t seem to me that Lisak’s study was even relevant to a place like Bowdoin.

It started to become clear that there is no existing research that can give us an accurate understanding of how often false accusations happen at colleges.

I started questioning other statistics I’ve heard. The study finding that one in five college women experience sexual assault or attempted sexual assault has been widely cited. The study only surveyed students two large public universities, however. Again, two large public universities cannot adequately represent the entire nation.

The study also included someone “rubbing up against you in a sexual way, even if it is over your clothes” in its definition of sexual assault. While this forced contact is undoubtedly unacceptable, it is clearly not what people imagine when they hear “sexual assault.” This data, too, did not seem applicable to our conversation.

I looked into the Center for Disease Control (CDC) report that recently found that one in five women in America (not just in college) experience rape or attempted rape. The study found that two million women, 1.6 percent of the female population, experienced rape or attempted rape in 2011. In a Department of Justice survey, 250,000 women reported experiencing rape or attempted rape. Confusing things further, the FBI found that only around 80,000 cases were actually brought to police.

Perhaps this discrepancy comes from the CDC’s language: it asked women if they had ever experienced specific sex acts while “drunk, high, drugged, or passed out and unable to consent.”

Critics have been quick to point out the word “or” prompts women to include drunk or high sexual encounters to which they consented. On the flipside, the range of encounters this language encompasses allows for the inclusion of instances that women might be hesitant to self-label as rape. Should we be including instances that the victim herself doesn’t consider rape? There doesn’t seem to be a clear answer.

The more I learn about these statistics, the more I realize how unhelpful they are. I don’t buy just 80,000 cases in a year, but I don’t buy two million either. It’s nearly impossible to come to a conclusion about rates of sexual assault when there is no widespread agreement on what sexual assault and consent are.

Ultimately, I don’t think we can apply this data to our everyday lives. My friends can’t prove their fears are valid and I can’t prove they aren’t. We’re all just going off our contradictory and irreconcilable gut feelings.

All I can say is this: I’m going to pay close attention to the sobriety of my sexual partners to make sure that they’re not anywhere close to “incapacitated.” If I do that, I seriously doubt I’m going to be accused of sexual assault.