The IWF Attack On Rape Statistics

In the National Review Online, and also on the IWF blog, IWF vice-prez Carrie Lukas critiques of Mary Koss’ groundbreaking study of rape prevalence. Lukas’ target is Koss’ finding that 1 in 4 college women has experienced either rape or attempted rape since age 14.

The one-in-four statistic… was derived from a survey of 3,000 college women in 1982. Researchers used three questions to determine if respondents had been raped: Have you had sexual intercourse when you didn’t want to because a man gave you alcohol or drugs? Have you had sexual intercourse when you didn’t want to because a man threatened or used some degree of physical force… to make you? And, have you had sexual acts…when you didn’t want to because a man threatened to use some degree of physical force… to make you?

Based on women’s responses, researchers concluded that 15 percent of women surveyed had been raped and 12 percent had experienced an attempted rape. Therefore, 27 percent of women … more than one in four … were either the victims of rape or attempted rape. This is the origin of the one-in-four statistic.

Yet other data from that same survey undercut its conclusion. While alcohol surely plays a part in many rape cases, the survey’s wording invites the label of rape victim to be applied to anyone who has ever drank too much, had a sexual encounter, and then regretted it later.

Yes, that’s a concern – out of context, I’ve always found Koss’ question about alcohol distressingly ambiguous.

However, it’s not enough to express concern. We should also ask, what does evidence say? Anti-feminists have been repeating this criticism of Koss’ survey for at least 15 years, but I’ve never seen one provide a speck of evidence that the question, in the context of a survey about rape and sexual coercion, is actually misunderstood by respondents to mean “have you ever had sex while drunk and regretted it in the morning?”

In fact, evidence shows that Lukas is wrong. Researchers Martin Schwartz and Molly Leggett tested the disputed question empirically back in 1999.1 They surveyed students with a modified version of Koss’ survey, which substituted this question for Koss’ original alcohol and drugs question:

Have you engaged in sexual intercourse when you didn’t want to but were so intoxicated under the influence of alcohol or drugs that you could not stop it or object?

If Lukas and other critics are correct to believe that Koss’ question creates a significant “false yes” problem, leading Koss to overestimate rape prevalence, then a significantly larger proportion of students would have answered “yes” to Koss’ original question than to Schwartz and Leggett’s rewritten version. So what actually happened? Rewriting the question made no difference at all – 17% percent of students surveyed by Schwartz and Leggett were found to have been raped, a number basically identical to Koss’ 15%. This proves that Lukas is wrong – Koss’ results are not caused by students saying “yes” because of morning-after regrets.

This result is unsurprising, because Koss and her co-researchers did extensive validity testing of the questions to make sure that they weren’t misunderstood. If a lot of students had misunderstood the question as referring to next-morning regrets, the question would have been rewritten early in the process. (So why didn’t Lukas mention the validity testing? For that matter, why didn’t she mention Schwartz and Leggett’s 1999 research?)

Lukas continues:

In addition, only 25 percent of the women whom researchers counted as being raped described the incident as rape themselves.

This misstates, in a subtle but very important way, what Koss’ study asked. 73% answered no to the question, “it was definitely rape” (emphasis added).

We have to consider context: we’re talking about young women, most of whom were raped by someone they knew (usually someone they were dating and had already been sexually fooling around with), who were in high school over 20 years ago, when discussions of date rape were extremely rare. It is any surprise that most of them weren’t positive that their experience was “definitely” rape?

In the real world, women who are raped – even in situations which anyone would call rape – are frequently, for whatever reason, not prepared to name what happened to them “rape,” let alone “definitely” rape. As Schwartz and Leggett noted, even among women who were physically forced or drugged into absolute helplessness – experiences that even the most determined anti-feminists will ruefully admit are rape – many or most refuse to label their experience “rape.”

What are the implications of deciding, as Lukas does, that if the victim doesn’t say it was ‘definitely’ rape, it’s not? Consider these statistics from Koss’ survey: 70% of the alleged rape victims in Koss’ study resisted by physically struggling with the man, and 84% tried to reason with him to no avail. The large majority reported having sex when they didn’t want to due to force or threat of force.

Lukas’ argument is that it doesn’t matter if the woman resisted physically, tried to reason with the man, and felt they had unwanted sex due to force or threat of force; rape isn’t defined by non-consent, it’s defined by whether or not the victim checks “yes” by the words “it was definitely rape.” Should anyone be comfortable with that logic? Should the law really be that even if someone physically holds down an unwilling woman and shoves his penis into her vagina by force, it can’t be rape if the victim, for any reason, doesn’t say it’s “definitely” rape? That must be what Lukas thinks, if she applies her logic consistently, but does it make any sense?

Lukas goes on:

The survey found that four in ten of the survey’s rape victims, and one in three victims of attempted rape, chose to have intercourse with their so-called attacker again.

This critique of Koss just restates the old “a woman who stays must not really have been abused” myth. It’s bullshit when said regarding battered women, and it’s bullshit when said regarding raped girls and women, too.

In Lukas’ fantasy world, of course a subsequent encounter – which may or may not be voluntary – proves that the earlier encounter wasn’t rape. The real world isn’t that tidy. It’s extremely common for victims of abuse to stay with their abuser for a while – certainly long enough for another sexual encounter.

Marcella at Abyss2Hope (before she started guest blogging here) addressed this, writing:

From personal experience I can speak to this paradox. My boyfriend didn’t fit the profile of a rapist as I’d been taught (a monster who snatches girls off the street) so even though what happened to me was rape, I couldn’t accept that he meant to treat me that way. I couldn’t accept that the guy who had been in my life nearly my whole life and who was one of my brothers’ best friends could be a rapist.

Looked at without understanding, people could think I decided to have intercourse with my boyfriend again. I did no such thing. It took a second rape (when I was still in shock from the first rape) before it began to sink in that the first time hadn’t been a fluke. He hadn’t mistaken the signals of non-consent.

Two rapes by the same person don’t cancel each other out or imply consent.

If you still don’t understand, think of it this way:

On the positive side of the scale I had 10 plus years of fun when this guy was around.
On the negative side of the scale I had less than 1 day of unimaginable pain and betrayal.

(I really recommend reading Marcella’s entire post).

Finally, Lukas concedes that “Another study…. found that one in eight American women … about 12 percent … had been victimized.” She makes it sound like this study stands alone. In fact, study after study after study – including major studies by the federal government – have found that between 10% and 18% of American women are raped at some point in their lives. These studies have used a variety of methods, worded the questions in various ways, and in some cases used extensive interviews to confirm that the questions were not misunderstood. There is no longer any legitimate argument over this matter; Koss was right to say that there is a great deal of “hidden rape” unmeasured by FBI and official statistics, and her anti-feminist critics were wrong to accuse her of deceit and exaggeration.

In social science, replication is considered the strongest evidence; if a finding is replicated by independent studies using various methods, it is considered strong. Koss’ findings, by this standard, are strong. This is a settled question. Rather than continuing to slandar Koss and distort her findings, the IWF should throw its political weight behind rape-prevention measures, such as anti-rape education aimed at middle schoolers.

* * *

In Marissa’s comments, Just Another Disenfranchised Father wrote:

However, I think that the point of Lukas’s article was not to suggest that going back to the purported rapist means that a rape has not taken place. I think she, and Christina Hoff Sommers, are pointing out the intrinsic inconsistencies of the survey which resulted in the 1 in 4 statistic and this makes that statistic suspect.

This makes no sense. If Lukas doesn’t believe that future sex encounters establish that all prior sex encounters were consensual, then where is the “intrinsic inconsistency”? The two things are inconsistant only if you believe that in a large majority of abusive relationships, the abused party leaves the abuser immediately after any case of serious abuse; but we know that’s not the case.

Comments for this post are open only to feminist and pro-feminist posters. Non-feminists may respond to the identical post at Creative Destruction.

  1. Schwartz, Martin D. and Molly S. Leggett (1999), “Bad Dates or Emotional Trauma? The Aftermath of Campus Sexual Assault,” Violence Against Women 5(3): 251-271. []
This entry posted in Anti-feminists and their pals, Mary Koss controversy. Bookmark the permalink. 

15 Responses to The IWF Attack On Rape Statistics

  1. Pingback:

  2. Pingback: Majikthise

  3. Pingback: feminist blogs

  4. Pingback: FeministBlogosphere

  5. Pingback: F-Words

  6. 6
    Rachel S. says:

    I would highly reccomend looking this survey from 2000. It has relatively similar findings. It is the basis for my rape lactures in my classes.

    There is some very finding things, especially when you look at the methodology.

  7. 7
    Ampersand says:

    That one’s my favorite, Rachel. (Yes, I have a favorite rape prevalence survey.)

  8. 8
    Sheelzebub says:

    These anti-feminists love to beat dead horses, don’t they? Or maybe it’s just their attempt to establish falsehoods as conventional wisdom–repeat KOSS IS WRONG! WRONG! WRONG I TELL YOU! enough times, and people will start to believe it.

  9. 9
    Rachel S. says:

    Yeah, a favorite rape prevalance survey. LOL!! I don’t know what that says about you Amp. LOL!!

    One of the interesting findings is how many of the women didn’t think it was rape…..When I tell my students about the questions they usally laugh, but they don’t realize that you can’t ask women if they were raped and capture many of the rapes out there.

  10. Thank you for an excellent post, Amp.

  11. 11
    Matt says:

    In a class I taught this semester we discussed when unwanted sex becomes rape. The main text used for discussion was a truly excellent paper by Sarah Conly- “Seduction, Rape, and Coercion”, from the journal _Ethics_ in 2004. It’s well worth reading. What was especially interesting to me, though, was that of the students who wrote papers on the subject, the men were much more sympathetic to a strict standard not only for what counts, legally, as rape but what sorts of behavior, while not being rape, is seriously morally wrong. It was a small group of students, but still it was a weird experience. (Let me recommend Conly’s paper one more time.)

  12. 12
    Polymath says:

    another important reason not to rely on self-reporting that “it was definitely rape” is that the standards of “what is rape” is (and ought to be) a legal one, not a subjective one. in fact, experiencing or witnessing abuse as a child is very likely to skew anyone’s “was that rape?”-detector. the tragedy of abusive families (beyond the suffering of the family members, of course) is exactly that: it makes the victims unable to discern future abusive situations from acceptable situations (“violence against me is normal because that’s how it was growing up”).

    if answering “yes” to Koss’s questions has a high correlation to having experienced legally defined rape (which amp is arguing it does), then the answer to “was it definitely rape?” is irrelevant, and any criticisms based on the answer to that question can’t hold up.

  13. 13
    Ismone says:


    Great post. I hadn’t heard that question defended before (it always gave me pause, and I read Hoff Summers’ book)–and your answer makes a lot of sense. I hadn’t even thought about validity testing.

    It’s also interesting that these antifeminists don’t come up with any stats. about how many women are really raped, or have any terminal point. Even if they could show that not all 15% that Mary Koss reported really were raped, who cares? So, an inflated number means we should what? Especially considering the fact that you pointed out–studies with (by their lights) better methodology have found the same result.

    The good old red herring defense.

  14. Pingback: Alas, a blog » Blog Archive » Ampersand’s 10 Best Of 2006

  15. Pingback: Alas, a blog » Blog Archive » NVAW Survey: Methodological Problems With The Question About Attempted Rape