I’m back from yet another whirlwind week of conferencing, since the annual Society of Vertebrate Paleontology meeting (this year in scenic Salt Lake City) just wrapped up last weekend. I’ll share some photos of the conference and welcome reception at the Utah Museum of Natural History soon, but today I’d like to talk a bit about who is giving talks at SVP and how we can increase speaker diversity. I hope you will share this with your colleagues and students!
This year at SVP, as in previous years, I was one of only a handful of female speakers talking about research on dinosaurs or other archosaurs. In order to delve into this phenomenon in greater detail, let’s just do a quick recap of how SVP structures its abstract submission process and conference schedule. SVP includes a couple of different formats for presentations:
- Technical Sessions (oral presentations): competitively scored abstracts submitted by society members; some abstracts are submitted for talks and are given posters instead because of the limited number of slots for talks relative to abstract submissions
- Poster Sessions: competitively scored abstracts submitted by society members; the vast majority of presentations at SVP are given in this format with about 150-200 posters presented each day.
- Podium Symposium (oral presentations): symposia topics are proposed about a year before the conference and presenters are invited by the symposium conveners; the abstracts are still competitively scored
- Poster Symposium: similar to the podium symposia but in poster format
- Romer Prize Session (oral presentations): senior grad students submit a special application in addition to their abstract in order to be considered for a talk slot in this session; highly competitive and requires extra work
When authors submit their abstracts for review for the conference, we’re given a couple of options for what format we prefer: consider the abstract for a talk but if a talk isn’t possible then a poster is ok; only consider for a poster; consider for a talk but if not then withdraw abstract. The abstracts are then reviewed by (I think) 5 reviewers, scored numerically, and the top-scoring abstracts get talk slots and other acceptable abstracts get posters.
It’s impossible to go to every talk at SVP given that there are four concurrent sessions each morning and afternoon, and invariably I don’t get to sit through all sessions because there are lots of other things occupying my attention at SVP. The talks are usually grouped by taxon, so similar groups of animals are covered in a morning or afternoon session. I usually manage to sit through the bulk of the dinosaur presentations and either keep notes on the presenters or recognize the names of the people presenting. How many women were first authors (and therefore, typically, the presenting speaker) in the dinosaur talks this year? Let’s take a look:
- Sauropodomorpha Technical Session: 0 women, 8 talk slots
- Theropoda Technical Session: 0 women, 10 talk slots
- Ornisthichia Technical Session: 2 women, 10 talk slots
I have to be honest, this is both pretty dire and also par for the course. And those two women in the ornithischian session? One of them was me. Over the last 5 years or so I’ve informally kept track of how many women presented in the dinosaur sessions and it has rarely (never?) been more than 3. Where are my female colleagues?
Let’s take a look at the breakdown for three of the Podium Symposia, one of which I attended in nearly its entirety (the molecular symposium); I checked the names of the other symposia by hand to see who presented:
- Molecular Symposium: 11 women, 16 talk slots
- Paleo-Evo-Devo Symposium: 5 women, 16 talk slots
- Endothermy Symposium: 0 women, 8 talk slots
Here’s where things start to get interesting: the endothermy and evo-devo symposia were convened by men [update: I goofed on the conveners for the endothermy symposium, which included 2 women; blame my late-night brain for missing this and my apologies to the conveners! See also M. Silcox’s comment at the end of this post], but the molecular symposium had a woman on the organizing team. I am not alone in noticing that a symposium with a female convener had more women presenters.
SVP abstracts are reviewed in a double-blind fashion, meaning the authors of the abstracts don’t know who their reviewers are, and the reviewers aren’t given the names of the authors for the abstract. In theory (and in practice), double-blind review helps reduce gender bias in the acceptance of scientific papers, so I would hope that we can safely rule out bias from the abstract review committee as a reason there aren’t as many women presenting talks at SVP. My working hypothesis for the past few years has been that the absence of women as oral presenters is because women are not opting to be considered for oral presentations at abstract submission time. If women don’t put themselves forward to give a talk, double-blind review won’t solve the problem. As such, we probably need more women to request oral presentations in order to increase the number of female speakers at SVP. I think the high proportion of female speakers in the molecular symposium and the low numbers in the technical sessions suggests that, broadly speaking (broadly! speaking! please don’t get upset at this generalization!), when women are invited to speak, we speak. But we don’t invite ourselves. And men need to get better at inviting us.
This hypothesis is supported by some recent discussions around trends in job applications, where women tend to apply to positions only when they match a high number of the listed qualifications, whereas the threshold number of qualifications is lower for men to submit an application. There’s also this example of an evolution conference where women were less likely to ask for the longer, 12-minute talk option vs the 5-minute talk slots. And don’t forget the link I mentioned before between women on the organizing committee and the number of women presenting.
While a bit discouraging, there are several concrete actions we can take (right now!) using this data:
- If you’re convening a symposium, invite more women to speak in your symposium. Make an effort to seek out female researchers on the symposium topic, especially people who maybe don’t often get up to speak. Here is a great blog post with concrete suggestions for increasing diversity in your symposium. Women in video gaming are also making good suggestions for how to increase diversity in what is currently a very, very skewed environment.
- If you’re a supervisor of female grad students or in any kind of mentoring capacity for female colleagues, encourage them to tick the box for oral presentations. Find ways to boost their confidence in public speaking, and teach them to write punchy, data-rich abstracts with strong narratives that are likely to score well with the abstract review committee. Make it really, really normal for everyone to request a talk. Take steps to avoid unconscious bias (we all have it!) that might influence your mentoring style with different groups of people.
- If you’re a woman, TICK THAT BOX THAT SAYS YOU WANT TO DO A TALK. TICK IT. DO IT. YOUR PROJECT IS GOOD ENOUGH FOR A TALK.
I cannot keep count of the number of times I have heard something along the lines of “Oh I’m not sure this project is ready for a whole talk” from my female colleagues. And while I agree that some topics (e.g. descriptions of new specimens) might be better suited to the discussion-rich format of posters, I also believe that you can find an interesting narrative for most projects that make them suited to the oral presentation format. I also cannot understate the importance of getting up in front of a crowd and being *noticed* for your work. Being noticed has benefits: you get feedback, you get collaborators, you get respect. And so, I just feel like I need to shout this from the rooftops: fellow ladies, your projects are good. Your projects are good enough for a whole talk. MAKE your projects good enough for a talk. Let the abstract review committee decide if your project isn’t the right fit for a talk. You will not get a talk slot if you don’t ask for one. I want to see more of you up on the stage with me when I’m invariably moderating the ornisthichian talks again in the future. Let’s make a promise to ourselves that next year you’ll tell yourself that your project is awesome and you’ll just tick that little box that says you want to do a talk.
I can’t invite each and every one of you to a symposium, but consider this a personal invitation from me to present in the technical sessions. There. You’re invited. See you in Calgary.
Thanks for sticking around all the way to the end of this post even though there was a catastrophic absence of fossils. Your reward is this Dyoplosaurus tail club, enjoy!
pseudoplocephalus 
Great article. I’m on a history-of-paleontology kick right now, and what’s really sticking out to me is the lack of information or even name recognition of female and non-Western paleontologists. I don’t feel like I’m getting the whole story. I realize there probably weren’t very many before the mid-20th Century, but even in the modern era, you yourself are one of the few female paleontologists I can name offhand.
Also, I’m sorry for the off-topic comment, but wow, that tail club! Or rather, those tendons! I didn’t realize the stiffened sections of ankylosaur tails were so extensive. How long is that specimen, and how does it compare to the average?
LikeLiked by 1 person
Ankylosaur questions are always welcome here! That tail club is about 1.3 m long, and tail clubs are usually between 1/3 to 1/2 the total length of the tail. Dyoplosaurus is a bit smaller than most specimens referred to Euoplocephalus and way smaller than Ankylosaurus, but still a pretty large animal. Notably, this specimen has a very small tail club knob – only about 15 cm wide – compared to the more typical 40cm found in lots of Euoplocephalus (the very largest tail club knobs approach and exceed 60 cm in width).
More info about Dyoplosaurus in this paper!:
https://www.researchgate.net/publication/262601230_A_Redescription_of_the_Ankylosaurid_Dinosaur_Dyoplosaurus_acutosquameus_Ornithischia_Ankylosauria_and_a_Revision_of_the_Genus
LikeLike
You’re job application comments are interesting. I definitely think that anyone looking for an academic job should be advised that it is the role of the search committee to reject them. They should not reject themselves. While I personally fell short of the “apply for everything that moves” strategy, I definitely was applying for multiple jobs each year that were a stretch. I will definitely advise anyone I meet at SVP to not disqualify themselves for things by not applying.
LikeLiked by 1 person
You’ll never get the jobs you don’t apply for!
LikeLike
One factual error here worth noting: the endothermy symposium had three conveners, two of whom were women (Jenn Botha and Colleen Farmer). I also know that over half the abstracts submitted to that symposium were killed by the program committee and some of the talks there were stuffed in by the program committee, not the conveners. My understanding is that this session had been intended to be much more gender balanced but it got majorly messed up by program committee.
I’ve heard similar complaints from other symposium conveners in past years.
LikeLike
Hey you’re absolutely right, thanks for pointing that out – I’ll blame my late-night brain for missing that Botha-Brink and Farmer were conveners for the endothermy symposium, and I’ll update the post. I’d be curious to know what the original slate of presenters had been, but your comment still reinforces my anecdotal point that when women organize symposia more women get invited to speak. A shame that it would up being so unbalanced!
LikeLike
Just to comment briefly on this as a member of program committee: invited symposium abstracts are subject to exactly the same review and scrutiny as any other abstract, and so have to meet exactly the same standards to make it into the program. Also, to reiterate a point Victoria made, program committee reviews are done blind – we genuinely have no idea who the authors are.
LikeLike
Yes, hopefully I didn’t give the impression that symposia abstracts don’t undergo review. My point is just that there might be a big difference between being personally approached to submit an abstract, vs. getting up the gumption to submit an abstract as part of the giant pool of submissions.
LikeLike
A question relevant to this: is the process of assigning talks to platform sessions done in a blind fashion? That is, after the abstracts are accepted are the program editors made aware of the identity of the presenter when they put together the sessions? Or are the schedule of what talks are presented and in what order done without knowledge of the ID of the presenter?
If the latter, than the distressing-masculine nature of the presenters of sessions that Victoria describes are an epiphenomenon of simply what got submitted and accepted. But if the former, than the program editors might take a more active role in shaping the sessions.
I don’t know which is the better approach, but something to consider.
LikeLiked by 1 person
For some insight into why, I’ll note that the heads of the program committee (one woman, one man) replaced symposium-intended abstracts which were not scored high enough to be talks with the highest-scoring talks that fit with the symposium’s theme. So this was, again, an issue of availability of high-scoring abstracts requesting talks–not a deliberate unbalancing.
LikeLike
I’m certain it wasn’t a deliberate unbalancing, but this is also interesting – if the symposium drew from a pool of not-specifically-invited abstracts, that means they were dipping into the pool that seems to contain fewer women-led abstracts (and thus the absence of female presenters makes a bit more sense). A shame that the symposium-intended talks weren’t scored high enough! Maybe we need to have a workshop on how to write effective abstracts?
LikeLike
This is not true. There were 11 abstracts submitted for the Endothermy symposium, only 1 of which was first authored by a female. That abstract was later withdrawn. Of the ten remaining, two were slotted in as posters based on their review scores (and one later withdrawn). The remaining 8 made up the symposium. We did not slot any talks into that symposium that weren’t submitted for the symposium. We did for the molecular symposium, in order to fill out the session with submissions that had reviewer scores high enough to merit a talk. In doing so, we actually increased the complement of female speakers.
We made the change this year of explicitly asking for a gender and career stage breakdown as part of symposium proposals. But fundamentally we work with what we’re sent, and make our decisions based on the quality of the work.
LikeLike
Hi Mary – Thanks very much for clarifying the endothermy symposium. Are there any plans to collect demographic information for the non-symposium abstracts? I know I’d be really interested to see how my hunch stacks up to actual data. And I agree, if there aren’t female-led abstracts submitted for talks, it’s going to be hard to fill the sessions with female presenters. Hopefully the end of my post gives some concrete suggestions for how society members can help improve the ratio in the future.
LikeLike
The anapsid, croc, and turtle podium session had similar numbers, with only 2 female-first authored abstracts.
LikeLiked by 1 person
Interesting, thanks for the extra info!
LikeLike
The proportion of women who attend SVP compared to those who actually get a talk is actually quite staggering…
LikeLike
I’d be really interested in what the breakdown is between the number of abstracts submitted for a talk by women vs. those accepted. Are we really submitting that many few abstracts? Or are our abstracts not being selected because they don’t score as high? And although I know that officially the abstract review is blind, I wonder how often people reviewing recognise the work they are reviewing and if there are any subconscious effects.
Small sample size, but I know that 3 women submitted for talks on pterosaurs this year, and we were all given posters. I know mine was scored just below the cutoff, which is fair, but I don’t know about the others. How are female abstract scoring compared to male?
Interesting questions… but definitely I think we need to submit more. If we submit more better abstracts, with the blind review system it should get better in theory…
LikeLiked by 1 person
You have definitely IDed a key issue here. Since the population being studied is the abstract submissions (rather than the SVP as a whole), we’d need to see the gender breakdown of submitted abstracts.
(Although if the breakdown of abstract submissions is radically different than the attendees as a whole, that would lead to another issue to address. So Many Issues!!)
LikeLike
One change we will be making for the coming year is to use first authors’ names, rather than initials, which will make it easier to track demographic data. We’re making this change largely at the request of woman members who are interested in looking at patterns through time.
With respect to what we can control, we have worked hard to ensure gender equality. We have continued the practice of double-blind reviewing. We have ensured that every reviewer group includes at least one woman, so every abstract is reviewed by at least one women. We have tried whenever possible to include women as moderators–although this was difficult this year because very few people volunteered to moderate, I think we were still pretty close to equal representation. What we can’t control are the fundamental demographics that underlie the abstract submissions. And what we won’t do is start making decisions based on anything other than quality.
LikeLiked by 2 people
Sounds like we’re pretty much in agreement here, but I’d like to reiterate that I don’t think there’s any issue at all with how the program committee and abstract reviewers are handling abstract submissions. I’m pretty sure the issue at hand is either 1) women aren’t writing abstracts that are getting scored highly or 2) that women aren’t submitting abstracts for talks, in favour of posters (my preferred hypothesis).
LikeLike
I am glad to see the double blind system in place and know that at least one women sees everything as well. What if you volunteer to moderate or serve on a committee, but never hear back from anyone? Who should you contact?
LikeLike
And with respect to your comment “Maybe we need to have a workshop on how to write effective abstracts?”, once I have finished being program chair (after Calgary), I would be happy to do a session on “how to write a well-scored abstract” for the Women in Science luncheon. I’ve now read literally thousands of abstracts, so I think I could maybe provide some useful tips. But obviously doing it while I’m still in the decision-making seat would be a conflict of interest.
LikeLike
That sounds like a great topic for a Women in Science luncheon!
LikeLike
That would be great Mary! I would love to see you do this in the future and I think it would be very helpful.
LikeLike
A project for those with more time than those who’ve posted so far (but which one could advise) might be a look at the gender ratios of a given topic (marine mammals; ungulates; non-avialian theropods; biomechanics; paleoecology; etc.) in the poster session, and compare these to the ratios of the same talks in the platforms. If the ratios are similar, then this probably points to a real gender discrepancy in interest/research in a given topic. If there is a distinction between ratios in poster vs. platform on the same topic, though, it might be evidence of (I would hope implicit) bias in choice over who gets to speak.
Another, longer, topic (a paper in sociology or in history of science studies or such) might be persistent patterns of gender proportions in the different subtopics.
LikeLike
I know I’m yattering on here, but this has got my brain storming. One possibility for data collection on demographics for the future (don’t think it can be automated for the past): tag demographic information to abstract submission numbers. The committee need not see this information at all in their selection process, but the data might be useable for future study. I know that we include name, nation of employment, and nature of position (tenured faculty, junior faculty, museum, etc.) in the registration for the Society. (I can’t recall if they included gender, honestly). Since we submit abstracts via the Society webpage, it might be possible to create a database linking this data to the abstract number. Then, once the fate of the different abstracts are decided (reject, poster, platform), someone would have a record of the results.
LikeLiked by 1 person
It’s pretty common for job application sites to collect voluntary demographic information for the purposes of understanding the pool of applicants and whether or not they are getting a diverse set of potential employees. This information is usually collected right before or right after submitting the ‘real’ part of the application (i.e. CV, cover letter, etc.), and I don’t think the information is linked to your application.
I think it would be cool if SVP did something similar with abstract submissions – right before the abstract is submitted, have a single page that collects basic demographic information (age, ethnicity, gender, nationality), but which isn’t linked at all to your abstract. The information can just be aggregated so that the society has a better understanding of the general population submitting abstracts, I don’t think it even needs to be searchable by person.
LikeLike
The SVP Program Chairs don’t select who gets a talk, we follow the recommendations of the program committee based on their review scores. Our main role is to group the abstracts into sessions and put together the program.
Here are some rough demographic numbers on this year’s SVP regular session submissions broken down by gender. These are quick tabulations, I didn’t track down the names when I was unsure, but they should be close (probably within a few percent).
Oral Preferred, Total: 103 of 382 = 27% were first authored by women
Oral Preferred, pre-cutoff (got talks): 70 of 250 = 28% were first authored by women
Oral Preferred, but shifted to posters based on score: 33 of 132 = 25% were first authored by women
Poster Preferred submissions: 147 of 327 = 45% were first authored by women
Total, SVP Regular Session: 250 of 709 = 35% first authored contributions by women.
These numbers are approximate, but as Victoria suspected, there are stark differences in submission patterns. It would take more work to break this down by taxonomic group, and we hope to be able to more easily get that data in the future.
LikeLiked by 1 person
It would be useful ultimately to also have this information for distinct career stages. I wonder about the influence of the ‘leaky pipeline’. Master’s students and PhD students are presumably less likely to request talks than postdocs and tenured faculty, but we also know that the leaky pipeline results in proportionately fewer women in those later career stages across most of science, presumably including palaeontology.
LikeLiked by 1 person
I guess what I think we need to know is whether women at a particular career stage are less likely to request talks than men of the same career stage. My gut feeling is that this difference will exist, but be most pronounced at PhD level. I’m pretty sure that all the women I know at postdoc level and above would have no issues with requesting talks, and I would guess Master’s and undergrad students across the board tend to prefer posters. If we had data we would however know for sure where mentoring efforts would best be concentrated.
LikeLiked by 1 person
Well, this is a very limited sample, but here are the career level demographics from the early synapsid session this year:
10 total talks (4 women, 6 men)
Career stage of women presenting: 1 masters student, 1 doctoral student, 2 postdocs
Career stage of men presenting: 1 postdoc, 2 junior faculty, 3 tenured faculty
So, somewhat higher-than-average representation of women in the session, but still pretty clear evidence of the leaky pipeline here.
(For reference on the total pool of applicants in this subfield, 4/11 of the synapsid posters were lead-authored by women: 3 doctoral students and 1 postdoc; I do not know how many of those initially asked for a talk.)
LikeLike
To address Victoria’s original comments about the dinosaur sessions, and since the numbers are lower than the total SVP averages, I did a quick tally on the Dinosauria taxon subcategory:
Oral Presentation Preferred (total): 11/69 = 16% of total submissions were first authored by women
Oral Presentation Preferred, For Theropoda: 1/26 = 4% of total submissions were first authored by women
Poster Preferred: 31/70 = 44% of submissions were first authored by women.
These are distressing numbers for talk submissions. They also suggest that there is potentially significant variability between areas of focus within the Society.
LikeLiked by 1 person
Is it also possible that we find new, more inclusive ways to share earth science research? Talks are easy to organize but I have doubts about their effectiveness, whether given by a man or a woman. I think we need more experimentation by organizers, trying things like knowledge cafés, birds-of-a-feather meetups, Wikipedia edit-a-thons, author ‘office hours’, and unconferences. I’ve been to conferences with these kinds of events (eg SciPy and Wikmania), and they are less white-male-dominated, and more energizing, than the geoscience conferences I attend. They are different communities too, so of course there are other factors, but it wouldn’t hurt to try.
LikeLike
How big are those meetings, though? And how much overlap in interests for individuals? When SVP became its own thing, it was entirely what would now be called an “unconference”. But when groups get big, structure is needed, and SVP is very much beyond that threshold. The primary purpose of this meeting is to allow as many people as possible to present their scientific research effectively (that last word is why lightning talks have been nixed whenever they’ve been brought up). If you start experimenting, you either have to cut a big chunk of that part of the program out (telling more people they don’t get to present when we’re already denying 2/3 of talk requests; dunno the rate for poster requests or talk requests that get posters instead), add another day (I have yet to hear of anyone who wants that), or accept that there will be conflicts (hopefully in such a way that does not require upping the already-high conference fees to offset more room rentals).
Some of what you suggest already does happen as informal or official luncheons and workshops. Organizers of the informal ones could stand to put out notices earlier and in a more visible way, but given the wide range of subjects and likely variable interest year-to-year, I suspect it would be better to let this stay mostly or entirely in control of people in those groups rather than the program committee.
I don’t know how much you’ve looked at STEM demographics data like NSF’s S&E Indicators reports, but the only field of science with a higher proportion of white males than geology is physics. Vert paleo less so given the high number of female students (but the leaky pipeline is definitely still a problem we have). It’s getting better, but still needs a lot of work. Our under-represented minority inclusion in the US section of the membership is still pretty crappy, though. So normalization to differences between fields definitely needs to happen before any conclusions can be drawn about whether or not the formats you suggest create a more inclusive environment or simply happen to occur in environments that are already more inclusive.
On top of all that, if we started doing more group discussions and fewer individual presentations, there’d be that whole pesky well-documented differential between the times the sexes spend talking in such environments, are perceived to have spent talking, and, most importantly, are listened to. Which could throw things in the exact opposite direction of what we want. If a misogynist is at my talk, they have to sit quietly and listen for twelve whole minutes (bwahahahah!) if they want to voice their opinion during the questions without looking the fool in front of everyone by asking something I already answered. If we’re in a group discussion they can do things like refuse to look me in the eye so they can pretend I’m not there, only talk to my brand new baby undergrad male companion even though the newbie keeps deferring to me and my newly-minted M.S. (then keep acting really surprised when I say something smart), or treat me like a child who needs her hand held and the simplest things explained even though my advisor has vociferously sung my praises as a Ph.D. student. Posters are, of course, going to be vulnerable to this sort of interaction regardless of how we present them.
My experience at SVP has thankfully been that these sorts of people are either few and far between or don’t reveal themselves as such very often because they know the vast majority of SVP-goers would call them out on their crap (the SVP community is awesome like that). But the possibility is still there because two of those people I mentioned are vertebrate paleontologists, one of whom I met at SVP, and I’m crap at picking up on subtlety so who knows how much less blatant sexism has avoided my notice.
LikeLiked by 1 person
Pingback: Fossil Friday Roundup: November 4, 2016 | PLOS Blogs Network
Adding another data point from Eugenia Gold (I’d missed her message when it came in a few weeks ago – my apologies, Eugenia, but thanks for sending this in!): in the bird session, there were 2 female presenters out of 10 total presentations.
LikeLike
Pingback: SVP in SLC | pseudoplocephalus
Pingback: I’m a #reallifescientist | Kristina Barclay
Pingback: Your Project is Good Enough for a Talk: Green Chile Edition | pseudoplocephalus
Pingback: You should apply for that: 10 tips for scholarship applications – Time Scavengers