Skip to content

1 in 4 women: How the latest sexual assault statistics were turned into click bait by the New York Times

by Brian D. Earp / (@briandavidearp)

* Note: this article was originally published at the Huffington Post.


As someone who has worked on college campuses to educate men and women about sexual assault and consent, I have seen the barriers to raising awareness and changing attitudes. Chief among them, in my experience, is a sense of skepticism–especially among college-aged men–that sexual assault is even all that dire of a problem to begin with.

“1 in 4? 1 in 5? Come on, it can’t be that high. That’s just feminist propaganda!”

A lot of the statistics that get thrown around in this area (they seem to think) have more to do with politics and ideology than with careful, dispassionate science. So they often wave away the issue of sexual assault–and won’t engage on issues like affirmative consent.

In my view, these are the men we really need to reach.

A new statistic

So enter the headline from last week’s New York Times coverage of the latest college campus sexual assault survey:

1 in 4 Women Experience Sex Assault on Campus.”

But that’s not what the survey showed. And you don’t have to read all 288 pages of the published report to figure this out (although I did that today just to be sure). The executive summary is all you need.

Here is what the authors of the survey–prepared on behalf of the Association of American Universities (AAU)–had to say in their introductory remarks:

[E]stimates such as “1 in 5” or “1 in 4” as a global rate [are] oversimplistic, if not misleading. None of the studies which generate estimates for specific IHEs [institutes of higher education] are nationally representative.

They go on to highlight that only 19.3 percent of students who were contacted actually responded to the survey, despite incentives–a low response rate for these kinds of surveys–and that even they were not likely to be representative of the student body within their own schools.

Specifically: “An analysis of … non-response bias found [that] estimates may be too high because non-victims may have been less likely to participate” (see Appendix 4 of the AAU report for an in-depth discussion of the analyses used).

None of this is buried in the fine print. In fact, the authors of the report (still in the executive summary) explicitly chastise news organizations for their misleading coverage of previous surveys:

[M]any news stories are focused on figures like “1 in 5” in reporting victimization. As the researchers who generated this number have repeatedly said, the 1 in 5 number is for a few IHEs and is not representative of anything outside of this frame. The wide variation of rates across IHEs in the present study emphasizes the significance of this caveat.

Another caveat has to do with definitions. “Sexual assault”–an incredibly loaded term–can mean a lot of different things in different contexts. In this survey, it means “nonconsensual sexual contact involving [either] sexual penetration [or] sexual touching.”

“Sexual touching” includes “kissing” as well as “rubbing against the other in a sexual way, even if the touching is over the other’s clothes.” (I’ll say more about this wording later on.)

What about “nonconsensual”? This means either that the act was physically forced, or that the person’s consent could not be obtained because they were “passed out, asleep or incapacitated due to drugs or alcohol.” But no definition of “incapacitated” is given, so it’s not clear how drunk (to use the example of alcohol) you have to be to meet this particular condition.

Interpretations could range pretty widely.

Are you a victim?

Finally, consider that the survey’s developers “specifically avoided” using the words ‘rape’ or ‘sexual assault’ in the questions they administered to the students. This was so that “respondents would use a set of uniform definitions when reporting on the types of events that were of interest.”

There are some sound methodological reasons for doing this. But as Ashe Schow has pointed out, it can also have the effect of eliciting much higher responses than would otherwise be obtainable, due to avoiding such harsh-sounding words. As a result, “many students who don’t [actually] view themselves as victims and [who] don’t believe they were sexually assaulted” could end up being counted as “victims of assault.”

There’s an obvious response here. “Well,” you might say, “they were sexually assaulted, and they are victims, even if that’s not how they feel.”

But let’s try a little perspective. Technical definitions on a questionnaire may not, in fact, line up with someone’s lived experience, and people should be free to reach their own self-understandings when it comes their status as victims (or non-victims). And some students may simply disagree with the definitions. Given, however, that the students were forced to choose between “yes” and “no” in response to the various questions–and were given no chance to explain or elaborate–these are not just trivial details.

A more accurate headline?

Now, much more could be said about caveats, but using just the information we have so far, we can see that a more accurate headline would look something like this:

Approximately 1 in 4 of 19% of a Non-Representative Sample of Women Who Responded to a Non-Representative Survey of 27 Colleges (Out of Roughly 5,000) Reported Experiencing Sexual Assault, Where “Sexual Assault” is Taken to Mean Anything from Being on the Receiving End of an Unsolicited Kiss to Forcible Penetration at Gunpoint, Regardless of the Particular Context 

Obviously, that’s too long to be practical – so let’s see if we can think of something shorter:

Latest Campus Sexual Assault Survey Paints Complicated Picture: Report Stresses Methodological Limitations 

That’s about right in terms of accuracy … but it wouldn’t sell very many papers.

Faced with this dilemma, the New York Times, a discourse-shaping publication with a massive readership, elected to go ahead with a headline that the authors of the very report they’re covering went out of their way to–repeatedly–emphasize was “misleading.”

And that’s putting it nicely. If you take into consideration the likelihood of inflated estimates due to non-response bias, a controversial definition of assault, and a relatively small, self-selecting sample, then the headline is simply false. (The actual content of the article isn’t much better. The New York Times author waits six paragraphs before mentioning the 19.3% response rate, and even then only in passing.)

Why this matters

So here is my problem with the New York Times coverage. In my view, even one assault is one assault too many, and what this survey does show is harrowing enough. There is no need to exaggerate the findings. Also sexual assault is serious. It isn’t something to be dumbed down into a simple factoid or turned into fodder for a superficial headline.

Even more to the point, there is a risk that “overselling” these kinds of findings will actually backfire in the long run, casting a shadow of cynicism over the efforts of those who are struggling–against not inconsiderable odds–to establish a culture (but hopefully not a bureaucracy!) of affirmative consent on college campuses.

The last thing our society needs is one more excuse to take rape statistics lightly.

One more thought

I said I’d come back to the issue of kissing. Remember that this counts, in the AAU survey, as “sexual touching” – and therefore (possibly) a form of sexual assault.

I expect that many people will be tempted to roll their eyes. Kissing? Sure, it might be unpleasant if you weren’t expecting it, but is it really in the same “category” as rape? Aren’t these broad definitions just being used to “inflate” college sexual assault statistics, when what we really care about is something more violent?

You might think that I’d agree with this view (based on my qualifications about the survey’s methodology, above). But I don’t think it’s actually that simple. Context matters. If the person who kisses you against your will–or after one too many drinks–is your professor, or someone who’s been harassing you all semester long, or a friend who’s now violating your trust, the emotional consequences could be pretty severe. It also depends a lot on who you are. Some people experience even “extreme” forms of assault and yet somehow manage to recover and move on with their lives. Others may be emotionally handicapped for decades.

So violating another person’s sexual autonomy, even if it feels like “only a little bit” to you, is ultimately a moral non-starter. It isn’t worth the risk.

Improving campus culture and sexual assault discourse

This doesn’t mean that every drunken pass at a party should be treated like a horrible crime. But it does mean that people of all genders, including men, women, trans people, queer people, and other gender-nonconformers (the latter should be highlighted because they are the group that reported the highest rates of sexual assault in the survey) need to come together in a spirit of good faith–really listen to each other–and try to promote a culture of basic respect when it comes to sexual relationships.

The goal, in other words, should be to massively reduce the incidence of sexual harm–not only on college campuses, but everywhere. But in order to do this, we need to be serious about how we estimate the prevalence of assault, frank about the limitations of what we know and don’t know, thoughtful about how we define our terms (including the range of behaviors we feel comfortable collapsing under one label), and painstakingly careful about how we communicate our findings to the public.

I know far too many people who have experienced sexual assault, running the gamutof offensive behaviors and contexts. I have seen first-hand the enormous toll that these kinds of violations can take on people’s lives. These individuals–and all of us–deserve a sober conversation about sexual harm, and the way it actually plays out on college campuses. We can do better than misleading sexual assault statistics turned into click-bait by the New York Times.


Cantor, D., Fisher, B. Chibnall, S. et al. (2015). Report on the AAU Campus Climate Survey on Sexual Assault and Sexual Misconduct. The Association of American Universities. Available here.

* Note that in this version of this article, compared to the original at the Huffington Post, a link to the AAU report has been added in reference to Appendix 4 (where non-response bias was estimated), so that readers may see those analyses for themselves.

Follow Brian D. Earp on Twitter by clicking here.

Share on

2 Comment on this post

  1. I think there is a major cultural shift in progress on sexual behaviour, and questioning the validity of individual surveys is not the way to approach it. It is a complex issue, but this quote illustrates the problem. It’s from a column by Ashe Schow, which Brian Earp refers to:

    While on paper forced kissing sounds bad, think of how this has been employed in movies without it appearing to be sexual assault. (Remember when Indiana Jones sexually assaulted someone? Neither do I.) Again, on paper this seems bad, and we can all imagine a scenario where a forced kiss is indeed sexual assault, but it seems absurd to assume that all of them are.

    Now anyone who is familiar with real existing feminism, knows that feminists will be enraged by that. To them forced kissing is rape, absolutely and without qualification. It is always rape, every time, and under all circumstances. The scene from the Indiana Jones film does indeed depict rape, and it was put there because Hollywood is run by men, and Hollywood deliberately promotes rape, which in turn shows that rape culture is all-pervasive.

    The tone of feminist criticism is absolutist. In turn, the criteria which feminism uses to define ‘consent’ are rigorous to the point of excluding it entirely. For feminists, rape and sexual assault by men begin long before they get to the forced kissing stage. Their definition of sexual harassment begins with mental state, even before anything is said or done.

    So it’s not useful to question the definitions used by any individual survey. The point is that a survey using feminist criteria will find, that almost all women who are in regular contact with men are semi-permanently harassed and sexual assaulted. (The Everyday Sexism project is a good source of examples). A feminist survey will find that over 90% of women have been raped. That perspective is, in itself, a social and political fact. The surveys themselves are almost meaningless: what goes in, is what comes out.

    So we must rethink our ideas about sexual behaviour, and that goes way beyond courses on sexual consent. When I say “we” I mean political elites and the academic world, and the media, who seem generally blind to what is happening.

    So what is happening? Looking at social media, I see a trend: an increasing number of young women are overwhelmed by the degree of sexualisation of modern western society. It is logical to seek an explanation, in the pervasiveness of porn and porn-derived images and attitudes. The current prevalence of porn has absolutely no historical parallels, and that’s probably true for the degree of sexualisation, and the emergence of hyper-sexualised cultural elements. Sociology has underestimated their impact, mainly because the issue is not taken seriously.

    The young women who are overwhelmed by sexualisation turn to feminism, which is also now far more accessible than it was in 1970’s and 1980’s America, where second-wave feminism emerged. Modern feminist subcultures would be impossible without social media. Feminist attitudes do have some impact outside feminism, and it is no surprise to see the initial maximum impact in universities, where there is a concentration of young articulate women.

    The result is series of feminist political demands on sexual behaviour, which amount to a demand for the desexualisation of modern society. We can see in the ‘consent culture’, which is now promoted at American universities, a prototype for a general social restructuring. One of the first priorities in understanding this, is to look at who the feminists are, at why they are disaffected, and why they choose this route to their own political utopia. The social sciences have so far failed to do that.

    I don’t think there are any real mysteries here. In general, feminist disaffection is easy to understand, relates to biological facts about sexuality, and is capable of rational political analysis. However, it does not help that sexuality is still, to a surprising extent, a taboo subject for rational conversation. Only after that taboo goes, will it be possible to consider the ethical issues. (I mentioned some of them in earlier comments about sex robots). We need to talk calmly about rape, sexual assault, sexual harassment, and the function of women.

  2. #BlackLivesMatter is still so much less important than Feminism! As long as ALL women are oppressed by patriarchy why do we even worry about a very narrow oppression example – just a single race?

Comments are closed.