And if you’re arguing that over 14% (some of the respondents in the Krebs study would be raped during the rest of their college career) of college women are raped while at college, but it costs too much to deal with college rape, you’ve moved from weak ground to quicksand.
One point to note is that most of the studies of college-aged women seem to ask if they have been victims since they were at college, but do not necessarily separate times they were actually ON campus/in college, from times they were a college student, but not at college. That is, some of the assaults may have taken place on vacations back home, at an off-campus job, on spring break in some place warm, or at a non-campus party by a non-college person. The study that came up with the 1 in 5 certainly did that by doubling the rate women reported while at college to cover the half of the year they weren’t at college. Not sure if that makes a difference, but it might. Which way do assaults outside the college zone get counted?
Well, isn’t the BJS study the one that asks women if they were raped in a face to face interview-like setting with a stranger? I can understand why it’s artificially low.
IMO, I would say that the Krebs study is a reputable attempt to get the numbers, with the notable exception of the inevitable response bias / truthfulness problems.
I’ll put my cards on the table, in terms of what I believe based on what I’ve read.
-
“20% of women will be sexually assaulted in college”. With the usual definition of sexual assault (including but not just rape), I have no problems believing it. In fact, I actually think this is low. I would have guessed 40-70%.
-
“6.1 out of 1000 college women were sexually assaulted in college (BJS report)”. Whether it’s a 4+ year rate or a snapshot rate, this is ridiculously low. I simply don’t believe it.
-
“20% of women will be raped in college.” - (Completed rapes, 4+ year rate) - Honestly, seems high to me. I’d guess lower. I believe the Krebs study implies about an 11% rate, correcting for a few factors. Using their definition, this is plausible to me.
To be honest, the biggest problem I have is the definition of rape (BTW - what Emma describes happening to her is a rape. Full stop.) Here’s the definition I would use. In addition to the usual graphic physical details, I’d ask the victims the following -
Hope you understand my intent. Too much typing today.
Alh - as someone who’s spent lots of time in all the stereotypical all male environments (but not prison!), I respectfully completely disagree. Whatever shifting occurs is quite small. I don’t want to get graphic, but the perpetrator would be at a lot of risk. I’m sure it is different in prison though.
Don’t draw too many conclusions from that BJS paper about college rape. The numbers are so small that they have to average over three years to even get something that looks remotely plausible, and even then the error bars are like 50% of the values. If you look at the raw data, the numbers are tiny and noisy.
Plus, the underlying data is bogus. It’s not only noisy, it’s woefully undercounted.
I’d want to class attempted rapes with completed rapes when counting. Other sexual assaults are also concerning, but IMO should draw lesser punishments. I’d say 11% might be roughly right for completed rapes plus attempted rapes, where I’d define rape as penetration of one of the three orifices without consent.
CF, I guess my argument is we don’t know…
If 11 percent of the women are raped in college, we have a very large reporting problem, disciplining problem, arresting problem, convicting problem…
A number of colleges have been taking campus climate surveys and publicizing the results. Here’s Cal Poly SLO’s:
http://campusclimate.calpoly.edu/include/CampusClimateFullReport.pdf
They say that 10% of the women respondents said they’d experienced “unwanted sexual contact,” which they defined as "forcible rape, use of drugs to incapacitate, forcible sodomy, gang rape, sexual assault, sexual assault with an object, and forcible fondling.”
We can think about correcting that number. It might be too low if students who are raped then leave the college-- they wouldn’t be sticking around to answer the survey. It needs to be adjusted upwards to allow for the fact that some respondents will be assaulted at Cal Poly but haven’t been assaulted yet. It could be too high if students who were sexually assaulted were more likely to respond to the survey. And notice it includes sexual assaults that are not rape. But it gives us a feel for the order of magnitude of the problem.
dstark - well, let the record show you’re the one who asked me. I don’t know how much formal training you’ve had in math. It sounds like either Dr. Fang / Cardinal Fang have had some. Anyway, I don’t dispute your common sense conclusion at all. I completely agree with it.
But here’s a pedagogical example for why a formal probabilistic analysis is complicated. This is in addition to the problems I’ve already mentioned, but I can do this in words rather than equations, so it might be easier to follow. Without some formal background you might not appreciate it or care.
Consider the Lisak / serial rapist issue. Let me concoct three extreme versions
-
The extreme serial rapist hypothesis™ - all rapes in a given zip codes are done by exactly one man. Everyone else is an angel. Suppose Adam comes forth, and he has indisputable proof that Paul assaulted him. In that case, it’s pretty clear that in any reasonable framework the odds that Paul is guilty of raping Emma go up dramatically.
-
The extreme anti-serial rapist hypothesis™ - all men are potential rapists. Whenever they’re alone with a woman, Beelzebub throws a die with thousands of sides. If it lands on snake eyes, he rapes her. In this case, Adam proving that Paul raped him tells you absolutely nothing about whether Paul raped Emma. It’s completely irrelevant.
-
The ultra-extreme anti-serial rapist hypothesis™ - no one commits multiple rapes. After committing a rape, a rapist gets it out of their system. Think Mr. Spock going through Pon Farr (sp) or Dr. Jekyll drinking his only shot of his potion. In this case, Adam proving that Paul raped him actually completely exonerates Paul from being guilty of raping Emma.
My point isn’t that any of these scenarios are true. My point is that whatever framework you are using to conclude that Adam coming forth must have some embedded assumption for what the Lisak /serial rapist study’s true results are. My scenarios show that you get radically different results depending on what the input to your model is. Since you didn’t consider it, there’s a big problem.
Realistically, the truth is probably somewhere between #1 and #2. It has to be considered if you want to squeeze a number out of the equations. Can’t avoid it.
BTW, the legal system seems to go with #2.
It’s one study. There are different ones that give different answers. You may be right about the statistic … in college, there are tons of single young men and women in close proximity, plus alcohol, My argument was with your choice of control group, not with the exact statistic. Anyway, we’ve beat this into the ground.
If a rough guess of women sexually assaulted at residential colleges is 10%, that’s WAY TOO HIGH and is a serious problem.
We seem to be assuming that the victims here are all women. Lately our area news seems full of stories of female high school teachers sexually assaulting boys.
Alsimon2, that is. a very good explanation.
I do not have the level of math you have. However, I had to deal with these situations in my career. It was very difficult at times. The 1987 crash happened one time. It is easy to say looking back that it could happen.
I was short call options in a stock and somebody made a takeover bid in that stock. The stock was up $10. I lost a lot of money. The guy who made the bid was living in a mental institution. This happened one time. Dow Jones printed the story without checking the source. How should I have priced the options for that?
I have dealt with these issues more than most people have. I think i am safe in saying this. 
The option models are not accurate on outliers. That’s ok. I understand and I understand everything you wrote.
When I went to school, I studied theories and formulas. The theories weren’t facts and many of the formulas don’t work. There were too many variables.
I had a marketing class where a formula was taught. I hated the formula. There were too many variables. 30 years later, I had a friend who was a CEO of an advertising agency. I asked him if he used the formula. He said, “I never heard of it. No. No way”. Lol
So? Alsimon2, I understand. I would love it if you did me this favor. Let’s pretend there is a 1 percent chance Emma is telling the truth. 50 percent chance Adam is telling the truth. 1 out of 500 chance there are 2 independent accusations.
There are no variables. That’s it. What is the chance the accused is guilty? What is the formula?
This is pretend.
Why are you concluding that Emma is telling the truth? It is a gut feel?
I sppreciate your taking time on this. It is helpful to me.
A yes, something I can understand…the theory of garbage in garbage out 
What is wrong with using the Clery Act numbers? Didn’t the feds just redefine what the colleges and universities need to report and isn’t this also administrated by the DOE? Seems like this is where the percent ought to come from and can be sliced an diced either by individual college or by types of colleges (so non-residential colleges could be eliminated). Since kids can chose the criminal system or the college system or both, the numbers would be accurate also. Even a student chose the criminal system only and never told the college, the local jurisdiction could report “back” to the college, which they might already do.
A new Cathy Young article on reaction to her DB story:
http://reason.com/archives/2015/02/09/flawed-narratives-perfect-victims-and-th
"The response from the rape-culture feminist camp has been to argue that there’s no “right” way to deal with sexual assault, generating a #TheresNoPerfectVictim Twitter hashtag. But it’s a straw (wo)man argument. Yes, of course victims deal with trauma in different, often startling ways. However, “no perfect victim” doesn’t mean that anything an alleged rape victim says or does, no matter how it defies common sense, reason and human experience, must be rationalized as “that’s what some victims do!” in deference to the commandment, “Believe the survivor.”
…Harris-Perry was complaining that all this talk about the need to get the other side of the story when it comes to rape accusations was terribly unfair because, for instance, when we do a story about ISIS, we don’t ask ISIS for comment. (Maybe that’s because ISIS actually puts out videos in which they saw people’s heads off or incinerate them. If there was a video of Nungesser choking, hitting and raping Sulkowicz, I can guarantee that no one would be going to him for his side of the story. Also, he’d be in prison.) Jelani Cobb, the panelist who did not get to say a word during my segment, gently tried to point out that checking the facts when it comes to accusations is important; as an example, he cited Joe McCarthy accusing people of being communists with no evidence and ruining their lives.
Harris-Perry’s response: “Yes, but he was just throwing out accusations! Women don’t do that!”"
There’s more in there, some I agree with, some not so much, but as usual I find it helpful to get all points of view.
She does not refer to the Jezebel article or reference the accusation from “Adam”.
Also wanted to post this on the Stanford swimmer case:
http://www.theguardian.com/society/2015/feb/08/stanford-swimmer-yik-yak-campus-rape
"After Turner’s arrest started making the rounds on the Fountain Hopper email, students didn’t take to the quad to talk about rape in public – they took their shock to their iPhones…
While the messages have vanished from the record, students described the Yik Yak conversations about the alleged rape after the Kappa Alpha party as a mix of collective horror, tasteless victim-blaming and outright misogyny – “Where all the unconscious bishes at?” read one. The posts contained little big-picture dialogue about prevention, awareness or even the two bicyclists who say they caught Turner before he could escape. Administrators and parents are nowhere to be seen or heard in the secretive campus debate.
Even when Stanford students did turn to public platforms such as Twitter, they were talking about the anonymous debate happening elsewhere.
“Another #Stanford student wrote on Yik Yak that it was ‘sexist’ to charge the perpetrator, as both parties were drunk,” tweeted computer science major Sasha Perigo."
I just appreciate Cathy Young’s measured responses so much. I’m following her now…she believes the way I do about alot of things, not just sexual assault on college campuses.
Fang: The Calpoly SLO report states that about 30% of respondents were not undergraduate students. It also says that 4.7% overall, 6% of undergraduates, and 8% of women were victims of assualt as they defined. I didn’t see the 10%? It also said that 80% happened off campus. I wonder if that includes frat houses? I wish these surveys subdivided out fondling from other forms of assault to really provide a better understanding of what those incidents involved.
Thanks for the welcome to the Board and some background for this discussion! I’m an attorney so I understand the different burdens of proof and why victims might want to seek redress in a college system and why colleges need to have ways to suspend or expel students who are dangerous to others. Having said that, adjudicating sexual assault charges is much more difficult for a college than assessing whether a student has plagiarized. These are hard cases to try even for experienced prosecutors. Here’s what I think we should want to happen in an ideal world when a student feels he/she has experienced sexually unwanted contact.
First, the student needs to know prior to that experience how to access services and support at her school as well as how to bring a complaint within both the college and legal setting.
Second, the system in place needs to have a reputation for being supportive of women so that they feel empowered to quickly seek the medical attention and legal support that they need.
Ideally, a student would receive immediate medical attention as needed, have access to counseling and other services through the school, and be supported by the state in filing criminal charges if deemed appropriate. If we truly want to see a decline in rape culture, criminal charges leading to trials and convictions are the best way to reach that end. The penalties are more commensurate with the crime than simply being expelled from college and the likelihood of it generating media attention (a la Vanderbilt), and thus possibly have a broader deterrent effect, is greater. I think it is very important that the college review process does not supplant the criminal justice system.
That does not mean that the student shouldn’t also seek redress on campus. The burden of proof and the evidentiary standards should be such that the college will be in the best position to protect its students - meaning to protect the rights of the accused to continue his education as well as the rights of the victim and all of the other students to be safe from sexual assault. For that reason, the burden of proof should be lower than the criminal standard and evidentiary standards need not be as stringent. For example, I think evidence that the accused may have committed other similar offenses should be admissible (which apparently was not permitted under Columbia’s system).
When women who have been sexually assaulted are as willing to press charges as those who have been robbed or endured a non sexual violent crime, that will dramatically help reduce the occurrence of rape on college campuses.
dstark - Honestly, I’d have to think about it for at least a few hours to come up with an approach that I was happy with that was also realistic. I’m not sure I have the time or that I’d even be successful.
Your point about the advertising formula is a good one. But please think about it. I’m actually making the exact same point you are with your advertising formula story. Getting a formula that works right is hard work. If you do a bad job then the answer is actually worse than your common sense. You only end up fooling yourself, or worse yet, it just gives you a weapon to try to browbeat others. That’s the only point I’ve been making.
But if you have time on your hands here’s a link to some related issues. It’s called the Prosecutor’s Fallacy.
http://en.wikipedia.org/wiki/Prosecutor%27s_fallacy
I’m sure Google can give you other references. What you’re asking about was first thought about over a hundred years ago. Check it out if you care to.
But the Prosecutor’s Fallacy stuff is actually the trivial stuff. There are lots of genuine issues, most of which I’ve forgotten or was never aware of. And that’s if you ignore all our concepts of due process and presumption of innocence, all of which introduce jurisprudence based overrides into the equations.
Here’s a simple example. Let’s say dstark believes only 2% of women lie, 98% are telling the truth. 98% > 50%, so why are we even bothering to have a hearing. Just get the rubber stamp out and grab an early lunch. Are you allowed to use the 98/2 fact? Why not? Where do you start from?
I understand your options background - I really do. Let’s just say that it’s really hard work to get the theories and the formulae to work well enough to beat the common sense that smart, experienced people bring to the table. Most people who try to do so are full of BS. But if you can do it then you can do really well. This algorithm http://en.wikipedia.org/wiki/PageRank made Larry Page a billionaire many times over.
The other problem with the Emma math probability is that she and the other two accusers talked which I would think adds a bias factor - they aren’t entirely separate or random. There are two common points - knowing Emma and knowing him.
I don’t know if anyone remembers the other Columbia case where the male accused student is now suing the school (quick refresher: no drinking, they were friends, she invited him to her suite bathroom and gave him a condom, but the next day she was concerned about her ex-boyfriend–who was also roommate of the accused—finding out. Months later she filed a complaint saying the sex was non-consensual. After accused was punished, accuser filed an appeal to reduce his punishment).
Anyway, I was struck by how similar the procedural complaints were on John Doe’s lawsuit and what the accusers against Paul Nungesser have said:
Natalie is accuser #2, Sara is now known to be Emma*
"But when Sara and Natalie were interviewed by a “neutral third party Title IX Investigator” named Jilleian Sessions-Stackhouse, they feared their stories weren’t being taken down in full: Sessions-Stackhouse didn’t use a recording device and took her notes by hand. "She would write things down that were abbreviations of what I said,” Natalie said. “Things that weren’t correct. It didn’t come out coherently. It didn’t sound like a strong case.”
Natalie went home for the summer and dropped contact with the investigator; her case was subsequently closed. Sara, who also noted Sessions-Stackhouse took incomplete notes—“I would be describing the position I was in when he raped me and her hand just wouldn’t move. She wouldn’t write it down. That’s important stuff,” she said—had her case overturned by a panel comprised of faculty members, and an appeal was later denied."
Unfortunately I can’t copy and paste, but if you go to page 16 #44 you can read the parts about Title IX investigator Jilleian Sessions-Stackhouse. Read through page 18, you’ll get the gist. Essentially, more incomplete and inaccurate notes, disregarding testimony, dismissing witnesses, etc.
As you can see, the incompetence and frustration goes both ways. Are these issues solely with this one person Sessions-Stackhouse being bad at her job, or is this more of a problem with the way these cases are treated on the whole?