Tuesday, April 25, 2017

Clancy: "Science doesn't have a sexual assault problem. LIFE has a sexual assault problem."

I just found a discussion of the SAFE13 study on the Breaking Bio Podcast. There are lots of things to discuss in it, but in this post I just want to highlight what Katie Hinde and Kate Clancy say about prevalence.

At 16:15, Morgan Jackson asks about this directly. Hinde first answers that they "can't really speak to prevalence" because of the self-selection problem. But "having said that" she goes on say that 71% of women and 40% of men in their sample did report experiences that "fit under the Equal Employment Office's umbrella categories for sexual harassment". What she says next is misleading. "Both men and women experienced a lot of marginalization, a lot of jokes about sex, a lot of jokes about cognitive sex differences, things like that." This leaves the impression that there were separate questions and "a lot" of women answered them in the affirmative. Actually, there was only one question* about this on the survey, and it asked about all of these (i.e., any of these) experiences, so we can only say that a lot (71%) experienced some of these things sometimes. Hinde makes it sound like the survey shows that a lot of people experience this sort of thing often. But, not only can they not "speak to prevalence" because of their sampling issues, this statement isn't even true of the sample.

Later, Clancy goes further, riding roughshod over the limitations of the study, albeit not to "belabor this prevalence point." She questions whether a completely controlled representative study of harassment could be done in practice, and then wonders whether this would actually give us any useful information. "We have absolute numbers of hundreds of women saying they were harassed and assaulted ... that's good enough for me." Again, this overstates what the survey actually shows. What they have is hundreds of women who say they have experienced anything from hearing a sexist joke (even just once) to being raped, and without actually specifying the exact behavior. And, in fact, as I've pointed out before, the survey can't distinguish between a grope and rape, nor an isolated remark from a campaign of abuse. So the "absolute" numbers are not quite what she makes them sound like they are. They're certainly not good enough for me.

What's interesting in putting Hinde's and Clancy's remarks together is the complete lack of any sense of proportion. Hinde doesn't care whether it's 75% or 50% or 20% or even 1% that get physically harassed. Even one is too many, she says. And Clancy, who actually does think she has everything she needs to assert high prevalence because they have hundreds of "absolute" responses, takes things to a new level by declaring that this isn't a problem in science but a problem in life. It's not that scientists are horrible people. It's just that they are, in fact, people, she says.

The problem with this should be obvious. Suppose the background risk of getting sexually assaulted (at your particular age) in "life" is either 5% or .5%. Now suppose that getting into fieldwork exposes you to either a 1% or a 10% chance of getting assaulted. These numbers do actually matter now. Getting them right is important. They matter as much as a man who is trying to choose between working on an oil rig and driving a truck in an oil field cares about his chances of being injured or killed. You can't just say that scientists are people and people rape and get raped. That's just not a serious statement, though it's about a very serious matter. The question is whether science is particularly dangerous. Not only do I see no evidence that it is; I see little evidence that the scientists who are studying the problem care.

The point is this. If an 18-year-old woman can reduce her odds of getting sexually assaulted by, say, half, simply by moving from the inner city to a college campus, then not only does academia not have a sexual assault problem, it appears to have found a solution to it. Likewise, if her risk of getting harassed as a 25 year-old is less if she chooses anthropological fieldwork over, say, business consulting or military service, then science (i.e., the culture of anthropology) appears have found an at least partial solution to the harassment problem that "life" poses. My guess is that universities at the very least offer a culture that both selects less rapists and keeps those that slip through better in check—better than other professional spheres, and better than society as a whole.**

Science, I suspect, is a safer place for women than "life". But Clancy and her colleagues appear hell bent on obscuring this fact.

*Update: for good order. Here are the only two questions that measure what sorts of experiences people had:

32. Have you ever personally experienced inappropriate or sexual remarks, comments about physical beauty, cognitive sex differences, or other jokes, at an anthropological field site?

39. Have you ever experienced physical sexual harassment, unwanted sexual contact, or sexual contact in which you could not or did not give consent or felt it would be unsafe to fight back or not give your consent at an anthropological field site?

If you answer "yes" to question 32 (as you must if you've heard a sexist joke, someone has said "typical female" about something you said, or, it appears, someone has complimented you on your legs ... once) you are counted as having been "harassed". If you answer yes to 39 (even if just to acknowledge that a colleague—or local stranger—once copped a feel at a party somewhere in the tropics) you have, according to the study, been "assaulted".

**An anonymous commenter, backed up by a correspondent whose opinion I respect, has noted that I seemed to be conflating rape, assault, and harassment in an earlier version of this paragraph. After thinking about it, I think I understand why they think so, and I've rewritten it to avoid this impression. It is, after all, something I'm accusing others of doing. Please note that I'm not claiming that business consultants experience more harassment than anthropological fieldworkers; I'm saying that it is interesting to know whether that is in fact the case. And I'm noting that Clancy and Hinde don't seem to be interested in such things.


Jonathan said...

Imagine this level of sloppy thinking in calculating the orbit of Jupiter's moons.

It is, in a sense, correct to say that Life has a sexual assault problem. That is in itself a huge step, because now we can calculate risks, as you suggest.

There is a fundamental mistake here, in that behavior potentially classified as sexual harassment is only such if it is repeated and pervasive. So hearing one joke, for example, is not sexual harassment under the legal definition.

Anonymous said...

Its hard to write about these things without making mistakes.

In your last paragraph you conflate sexual harassment, sexual assault, and rape in your logical inference. They might be correlated in a way that matters. I don't know.

It seems your evidence is as anecdotal as theirs. However, because of that, you both are making the point that the prevalence numbers are unknown and perhaps unknowable.

Allow me to offer the possibility that even approximate numbers don't matter as much as how each person FEELS about them. We might all contemplate that for a while.

I think you and I agree that in that regard, a lot of disingenuous adversity between activists on both sides is generating hostile feelings between even generally well-meaning colleagues. Perhaps the complementary approach could help: disingenuous admiration ... might generate good feeling ....

Anonymous said...

Jonathan, you have opened a new Pandora's box of possibilities for Title IX.

You wrote, "So hearing one joke, for example, is not sexual harassment ..." and I think, oh, wow, I had thought it could only be in the TELLING of the joke that one could be accused of sexual harassment, but I was too restrained. It contributes to the hostile work environment to even LISTEN to a joke. I began this as a parody, but now I remember there was a grad student who got Title IX'd for laughing at the juxtaposition of "juicy" and "girl scout" in an off-campus game. But that's in the telling of a cisgendered, hetero female feminist, so it might not be a fair retelling of events. Perhaps the student just THOUGHT about laughing and it all got blown out of proportion that the student had ACTUALLY laughed. No, wait, it should suffice for someone there to imagine that the student was thinking about laughing. No, silly me, it should suffice for someone who WASN'T there to imagine that the student was thinking of laughing. Oh snap - I am such a nit-picky lawyer today, demanding such constraints on witness testimony! It should suffice for someone who wasn't there to imagine that someone ELSE who wasn't there might be offended by the possibility that the grad student was thinking about laughing.

THERE. I think that's correct now. Given that there are two someone's against one grad student, the preponderance of evidence standard is met and we can move to the sentencing phase. (It doesn't matter that one of the someone's is imaginary.)

The latter (generally speaking, not specifically in the details) is approximately what got Laura Kipnis started in her Title IX inquisition: supposedly it wasn't that her accusers themselves were offended by her original article in the Chronicle of Higher Ed - they (bless them) were concerned that some OTHER students MIGHT be offended.


end of sarcastic response.

beginning of non-sarcastic response.

You make a good point about the limitation of the single infraction not meeting the pervasive standard. I know a common interpretation of the policy of Title IX is that a single event CAN be harassment if it is sufficiently egregious. In principle it all comes back to the necessary fiction of a "reasonable person" but in practice it depends on the whims and biases of the Title IX officer (or some other mid-level person, such as a Dept chairperson) and the political calculations of them and eventually the leaders of any given University.

Thomas said...

@Jonathan: Indeed, the question is "Have you ever personally experienced inappropriate or sexual remarks, comments about physical beauty, cognitive sex differences, or other jokes, at an anthropological field site?" The only evidence that Hinde has for saying that "men and women experienced a lot of marginalization, a lot of jokes about sex, a lot of jokes about cognitive sex differences, things like that," is that hundreds of people answered "yes" to that question.

@Anonymous: I agree that it's hard, but the "mistake" you point out was a deliberate economy. I passed from the risk of being raped at 18 to the risk of being harassed at 25. I wasn't comparing these two things, and therefore not conflating them. I was merely saying that rates of rape, harassment and assault are all measurable quantities and can, in each case, (and for each age cohort) be meaningfully compared across professional spheres.

(I'm not making excuses, but this is at the end of the day "just a blog post", an attempt to jot down my reactions to the video, so I'll grant that I could have avoided your misunderstanding with a little more effort, of course.)

I am very definitely saying that the prevalence of rape, assault and harassment are knowable facts. There is an "absolute" number of rapes in anthropological field sites every year, there is also an absolute number rapes in "life". Done right, the approximations of these numbers can meaningfully be compared. But this study just wasn't designed to even estimate the first number or allow for comparisons to the second. It simply doesn't count or measure anything "real".

So I guess it's quite apt to say what counts is how we "feel". I feel decidedly underwhelmed by this study. I can't even feign admiration for it. Sorry.

Anonymous said...

Maybe this question is a control?

32. Have you ever personally experienced inappropriate or sexual remarks, comments about physical beauty, cognitive sex differences, or other jokes, at an anthropological field site?

Because nearly everyone, if reading the question, will answer "yes, I personally experienced ... other jokes ... at a field site." There's nothing about the jokes being considered inappropriate or sexual, as written. Maybe the reader is supposed to read it as "or jokes of a sexual nature"?

So maybe it's like a lie detector test, where they ask some baseline questions ("what's your name? Your mother's name?") to calibrate your physiological response.

Thomas said...

They don't treat it as a control. If it's a lie detector test then 29% women and 60% of men are liars.

I think people read it as, "Have you ever experienced sexism directed at your person?" Even this, I guess, should probably be answered yes by almost everyone. Who hasn't experienced at least a little sexism? After all, these same researchers will tell you that everyone IS a little sexist (i.e., "biased") so it would be surprising if everyone did at times experience some.

So, answering yes probably depends on some subjective sense of "seriousness". As a "yes/no" proxy for "have you been sexually harassed?" it's pretty bad.

It reminds me of Andrew Gelman's image of a researcher trying to determine the weight of a feather in the pouch of a kangaroo jumping up and down on a bathroom scale.

Presskorn said...

Might be of interest to you: https://www.theguardian.com/world/2017/apr/02/unwanted-advances-on-campus-us-university-professor-laura-kipnis-interview

Anonymous said...


You might compare the Marcy-Ballard interaction with that of the unnamed astronomy professor and Kipnis' mother, as told by Kipnis in the book reviewed in the links by Presskorn or Anonymous. It does make you wonder, "Is there something about astronomers?" Or would that be profiling or stereotyping?