Tuesday, July 25, 2017


Here's question 7 of the CSWA Workplace Climate Survey:

In your current position, how often have you been VERBALLY harassed because of the following characteristics?

{Choices include: Often, Sometimes, Rarely, Never}
Race or Ethnicity
Physical Disability Status
Mental Disability Status
Sexual Orientation
Gender Identity(Cisgender or Transgender)
Gender (Female, Male, or Non-binary)
Religion or Lack Thereof

Question 9 has the same form, with "PHYSICALLY" replacing "VERBALLY". Now, here's hypothesis 1:

Female respondents will report more verbal and physical harassment than men.

This sentence appears in the first paragraph of the results section:

Thirty-nine percent of respondents report experiencing verbal harassment at their current position, and 9% report experiencing physical harassment.

And this one appears in their support for hypothesis 1:

Women were also significantly more likely than men to report that they experienced both verbal and physical harassment because of their gender.

I have underlined that phrase because it draws attention to the glaring absence in the questionnaire of a "characteristic" that is likely to have been focus of harassment directed at white male astronomers. Indeed, though the paper doesn't tell us this (we know it only from Christina Richey's preliminary presentations at DPS in 2015 and AAS in 2016), race and gender account for the great majority of characteristics that people felt they had been harassed for. In the case of verbal harassment, they account for 65% of the reports. In the case of physical harassment they account for more than 80%.

For obvious reasons white male astronomers are not likely to report being victims of harassment because of their race or gender. If they are also straight, cisgendered, protestant, able-bodied and do not suffer from mental illness, they would seem to have no way to report their experiences on the survey. And yet, surely, they might experience harassment. Most commonly, they will experience verbal (and at times physical) harassment by professional rivals with whom they are competing for publication, promotion and research funding. This basis for harassment has been completely excluded from the Workplace Climate Survey. The questionnaire did not even provide a generic "other" characteristic in which to report harassment.

Now, if the hypotheses tested had confined themselves to race- and gender-based forms of harassment, this wouldn't be such a big problem. Except that the conclusion that women and people of color experience more gender- and race-based harassment than white men is a bit underwhelming. But Clancy et al. claim to have found support for hypothesis 1, and they are promoting the result widely as suggesting that women, and women of color, experience more harassment than men full stop. As it turns out, this conclusion emerges from a measurement instrument that excluded the most typical experiences of harassment among white men.

I'm going to take some time to think about the consequences of this. But my initial reaction is that it completely undermines the validity of the survey, given the hypotheses it claims to be testing. Also, it raises the interesting question of whether women and minorities experience harassment based on professional rivalry (they would also not be able to report it). And that, finally, raises a question that concerns me greatly: is it possible that women and minorities are getting the axis of their harassment wrong? Is it possible that they experience harassment that is really grounded in ordinary competition as grounded in racism and sexism? If so, surveys like this are distracting them from the fight they should be fighting; and this, ultimately, will be to their disadvantage.

Comments are welcome, as always.


Anonymous said...

What completely undermines the validity of the survey is that the authors don't actually verify the identities of survey respondents, or their experiences. So anyone could claim to be a woman and/or minority who experienced verbal and/or physical harassment.

Thomas said...

I'm not quite that skeptical about the data. The survey was promoted in relevant places, so I believe that mainly members of the astronomy community took part. I don't think very many people lied about race or gender. And they did make some effort (using IP addresses) to ensure that people didn't respond twice. I would like to see a survey in which every attendee at AAS was given a unique (but anonymous) key, but that's wishful thinking.

As to whether respondents reported harassment that didn't happen: I want to stress that, given the sampling bias, the results are not really very surprising or alarming. Although 39% reported verbal harassment, less than 13% appears to have reported it occurring more than "rarely", and less than 2% seem to have reported it happening often. That's about seven people in a sample with a strong self-selection bias towards people who have something to report, and an (intentional) oversampling of women.

In the most extreme self-selection scenario we can imagine, everyone in the astronomy community who feels they're often harassed would have reported. In that case, .07% of astronomers experience verbal harassment often. I estimate the upper bound at 60 people who experience verbal harassment often (2% of female astronomers); in that case, about .6% experience verbal harassment often.

I don't think these numbers are so outrageously high that we need to imagine fraud on the surveys to explain them.

Anonymous said...

I would agree there's not necessarily a reason to suspect anyone lied about their identity. But from a scientific point of view, if your conclusions are based on an individuals identity, you should at least make some effort to verify that.

Personally, as an astronomer I'm not concerned at all by the results of the survey. As you pointed out, it showed the overwhelming majority of people hear or experience harassment either never or rarely. And I suspect the numbers given are higher than the actual rate, since those who have been harassed are presumably more likely to compete such a survey.

Anonymous said...

Thomas, you wrote:

For obvious reasons white male astronomers are not likely to report being victims of harassment because of their race or gender. If they are also straight, cisgendered, protestant, able-bodied and do not suffer from mental illness, they would seem to have no way to report their experiences on the survey.

I am curious what you think are those obvious reasons?

If I am dis-invited to a professional development workshop hosted by my employer because I am male, then wouldn't I validly respond to such a survey with "yes, I have been verbally harassed because of my sex"? I am confusing harassment with sex discrimination, but if it is pervasive, isn't that harassment?

Or do you mean that if a female astronomer says she'd rather not work with any men, then each male colleague that heard the remark would "take it like a man" and not report any discrimination, not even to an anonymous survey?

Or do you mean that white-hetero-cis-males dominate the field so much as a fraction of the population of professors that their white-hetero-cis-male colleagues or students will not ascribe any mistreatment from them to race or gender because they are the same?

Thomas said...

I mean a combination of those things, but mainly the last one. To take the simple example, in a male-dominated field males will rarely feel discriminated against on the basis of gender. It may happen, of course, (and much as you describe) but it will be rare (and, for most males, never). Any verbal harassment they will experience will be attributed to, say, status (a professor dressing down a student) or rivalry (a PI on a competing grant proposal gets drunk and abusive at a conference). This sort of harassment may be relatively frequent. Indeed, it may be as common as the gendered-harassment felt by women. And that brings me to the kicker: those women may be misunderstanding the reason they are being harassed. Even if a rival resorts to gendered slurs, there may be no sexism, as it were, "behind" the harassment. The reason she is being harassed is that she is a contender for the same position or grant or discovery. Ordinary human stuff.

My point isn't that there is none of the anti-male harassment you describe in science. There may well be. My point is that there is not likely to be a comparable amount of it, i.e., compared to the amount of gendered harassment women face, men face much less gendered harassment. The survey excluded the forms of harassment most men are likely to face if they face any at all. So the gender comparison is specious.

Anonymous said...

Thank you for clarifying. With that, if I understand you, then you think the survey should have included professional rivalry, as a control, as a point of comparison to judge the other forms of harassment?

So if it had a row labeled, "based upon professional rivalry?" and 0% said "never" and 80% said "rarely" and 15% said "sometimes" and 5% said "often" that would be helpful to interpreting the other results?

But I still don't understand why you feel the "gender comparison is specious." I think it is the main point of the survey: do male or female respondents to the survey experience (report) gendered harassment in significantly different amounts?

If a questioner of a speaker prefaced a question with a sexist attempt at humor, "Ironically enough, the subject matter of brown dwarfs is studied by a team of three unusually tall white men..." then I could imagine that those three might remember that comment and report it as gender-and-race-based harassment on the CSWA survey. If the questioner is male, the three may interpret the intent differently than if the questioner is female.

Maybe THAT is your point? That because astronomy is male-dominated by numbers of researchers, a female speaker is more likely to experience a male questioner (of the sort above) than a male speaker is to experience a female questioner? And under the hypothesis that only when the genders of speaker and questioner are not the same is ill intent impugned, that sets up a situation where females will report more gender-based harassment?

Can we make a simple physical model? Imagine a bowl full of M&M candies with two colors: green and red. There are four times as many green as red, but they are well mixed and identical in ALL RESPECTS except for the color. Then you take a survey and send it preferentially (but not exclusively) to the rare red candies, asking, "Are you experiencing inappropriate physical contact with a candy?" Then imagine both colors of candies equally feel that a necessary (but not sufficient) condition of *inappropriate* physical contact is that it be between candies of a different color. Assume idealized circumstances: everything is equal and response rates are not high but are random and the same for the two colors of candies. What will the result of the survey be? Will there be a difference in the reporting from red and green, aside from the intended disproportionate fraction of red candies in the set of respondents? I think there will be, but I have not worked out the statistics properly, and sometimes the correct answer is counter intuitive. If I make a limiting case of the ratio being very large, say 100 times as many green as red, then it's easier to imagine that nearly every candy (red or green) is surrounded entirely by green candies. So that skews the imagined survey results in a way that would have to be carefully corrected for, even in this idealized example. Even if the proportion of perpetrators is independent of color (red or green), then a survey will (I think) report that proportion from red victims but essentially zero green victims. Take the simpler idealized case of surveying ALL candies. Every red candy has the potential to be next to a green perpetrator, but a very large fraction of green candies cannot be next to a red perpetrator because they (the green candies) are surrounded by their fellow green candies, and by construction green cannot harass green (or red...red). I acknowledge that in the real world green can harass green and red can harass red, but in that same real world, often mild "harassment" is not interpret as such in those cases.

Maybe the latter (long winded) example was what you were thinking?

Thomas said...

The question "Do male or female respondents to the survey experience (report) gendered harassment in significantly different amounts?" seems to me to have an obvious and trivial answer: women will experience more gendered harassment. This does not mean that astronomy is "hostile" to women because men may experience as much (or more) harassment of another kind.

The important question is: Do male or female astronomers report harassment (of any kind) in significantly different amounts? Also, do astronomers of either sex report harassment of a significant amounts at all?

I think your example sounds right. The same behavior may be interpreted as gendered harassment, racial harassment or no harassment at all, depending whether it cross gender lines, race lines, or none at all. "Where are you from?" "You look hot today." "You need to get laid." These statement can be felt (and of course intended) as harassment, but are less likely to be taken this way when they pass between members of the same sex or same race. (Consider that last one said by a man to woman vs. said by one woman to another.) White males, meanwhile, will rarely be "positioned" at the "business end" of a remark that could be taken as race or gender harassment. But they can easily find themselves at the pointy end of a rival's sarcasm.

Anonymous said...

Ok, maybe you are being logical but I think it matters whether the harassment is of a Title IX protected kind or whether it is of another sort.

A rival's sarcasm is fair, isn't it?

As an example of such sarcasm, I will attempt to lampoon your comment. Consider males and females in combat roles in the military. A woman can argue it is sexist discrimination that she is not allowed to be in the infantry front lines solely because she is a woman, even if she is as able bodied as a qualified man by all objective measures. Her opportunity is curtailed solely because she is female. That is unfair to her BUT at the same time, that male experiences a more hostile work environment than her: he is face to face with the enemy while she is safely operating a drone five thousand miles from his location. The unfairness is wrong; the hostility is an intrinsic part of the job.

Thomas said...

Imagine a mean person who is always trying to gain an advantage and make others feel worthless. A bully. Let's make him a white man. When he's being mean to a woman he plays on gender stereotypes, when he's being mean to a person of color her plays on race. He actively deploys stereotype threat, that is, to put his competitors off their game. In the CSWA survey, his behavior would register has gender and race based harassment. It will be detected as sexism and racism.

But I think this is a mistake. When he is bullying a shorter white man he plays on height. When he's bullying a physical equal or superior he attacks their intelligence or style of dress. He also keeps track of embarrassing stories to use at key moments. He remembers people mistakes (in their science for example) and he spreads lies when there aren't enough mistakes to speak of.

The point is that the great majority of his behavior, directed at his white male colleagues (because there are more of them to abuse), will not be registered in the survey. He's not a sexist or a racist, actually (he only uses the tropes to his advantage), but he is an asshole. If we now imagine this person as a woman the content of her harassment might change but she's still an asshole. The survey will register her hostility differently--it will register less of it. Especially, of course, if the survey intentionally oversamples women.

It is precisely because it only looks at TIX-governed behavior that it arrives at the conclusions it does. It's like a study that only looks at the harms of illegal drugs. The real dangers (tobacco and alcohol) will be hidden.

Anonymous said...

That's interesting and an important interpretation. I will call it the Asshole Hypothesis. I believe I have known some of these. Even if it is entirely incidental, the asshole's racist/sexist harassment is wrong and should be changed or eliminated. But your point of how it might have biased the survey is interesting to consider. Also, the survey did not try to classify the perpetrators but only the victims.

The following is a variant on my red/green candy model earlier in this thread.

I wonder also about the number of interaction hours each population experiences. If there are four times as many men as women in the set of all astronomers, and if they interact without regard to gender, then the women will preferentially have four times as many F-M interaction hours as F-F hours. The men will have four times as many M-M interaction hours as M-F hours. If sexist comments are expressed the same by men and women (say, one in a thousand sentences uttered by one in 10 astronomers (of either gender) are sexist) but sensed and reported as such only when the interaction is with a person of the opposite gender, then females will report four times as many sexist interactions as males per unit time per unit listener. (This simplistic model treats two-person communication as the dominant form of interaction.)

Hontas Farmer calls this "a numbers game" and wrote of it Jan 22, 2016:


I think Farmer makes some errors, but I think the overall point is a valid one even if it is not legitimately a good interpretation of Clancy's results.

Anonymous said...

I should have written "then EACH woman will preferentially have four times as many ..."

I hope it was clear anyway.