Tuesday, September 11, 2012

Q-Bio conference in Hawaii, bring your surfboard & your Y chromosome b/c they don't take a XX

Wow.  Just wow.  And not in a good way.  Just got an email invitation to a meeting.  The meeting is
"THE FIRST ANNUAL WINTER Q-BIO MEETING: Quantitative Biology on the Hawaiian Islands. February 18-21, 2013."  
Well, I mean - who wouldn't want to go to Hawaii for a meeting.  And a meeting that 
"brings together scientists and engineers who are interested in all areas of q-bio."  
"Each year, the meeting will rotate on the Hawaiian Islands with a different thematic focus within q-bio."
So I could go to Hawaii each year.  Cool.  And 
"The focus for the meeting this year will be Synthetic Biology, with about half of the invited speakers chosen as renowned experts in this area."  
I like synthetic biology and, well, sometimes I like experts, so still good

But then, OMG, then, the confirmed speaker list and the conference organizers.
  1. Jim Collins, Boston University
  2. Johan Elf, Uppsala University
  3. Michael Elowitz, California Institute of Technology
  4. Timothy Elston, UNC Chapel Hill School of Medicine
  5. James E. Ferrell, Stanford University 
  6. Martin Fussenegger, ETH Zurich
  7. Leon Glass, McGill University
  8. Terry Hwa, University of California, San Diego
  9. Roy Kishony, Harvard Medical School
  10. Galit Lahav, Harvard University
  11. Andre Levchenko, Johns Hopkins University
  12. Wendell Lim, University of California, San Francisco
  13. Andy Oates, The Max Planck Institute, Dresden
  14. Bernhard Palsson, University of California, San Diego
  15. Gurol Suel, UT Southwestern Medical Center
  16. Chao Tang, Peking University
  17. John Tyson, Virginia Tech
  18. Craig Venter, The J. Craig Venter Institute
  19. Chris Voigt, Massachusetts Institute of Technology
  20. Ned S. Wingreen, Princeton University  
  1. Bill Ditto, University of Hawaii 
  2. Jeff Hasty, UC San Diego 
  3. Bill Hlavacek, University of New Mexico
  4. Alex Hoffmann, UC San Diego
  5. Brian Munsky, New Mexico Consortium 
  6. Lev Tsimring, UC San Diego 
That is a 25:1 ratio.  Pathetic.  Embarrassing.  The sponsors - UC San Diego's Division of Biological Sciences and BioCircuits Institute, San Diego Center for Systems Biology, the University of Hawaii and the Office of Naval Research - should all be ashamed.

For other posts on this topic see

UPDATE - I have now submitted an abstract to the meeting.  The abstract I submitted is available here and posted below

The probability of having one out of twenty six participants at a scientific meeting be female

A quantitative analysis of gender bias in quantitative biology meetings 
Jonathan A. Eisen
University of California, Davis

(Note - new title suggested by John Hogenesch)

Scientific conferences have key participants which I define to be the speakers and the organizers. Such key participants can be divided into two main classes based on gender: male and female, which I denote here as M and F, respectively (I realize there are other gender classes and I regretfully am not including them here). The number of key participants (which I denote as KP) for conferences varies significantly. For this analysis I focused on meetings with KP = 26. This value was selected for multiple reasons, including (a) that it is the number of letters in the English alphabet (b) that its factors include the number 13 which I like, and (3) because in email announcements for this meeting KP= 26. I sought to answer a relatively simple question - what is the probability that, for a meeting with KP=26, that F = 1. I chose this because this seemed extreme and because F=1 in the email announcements for this meeting. Using the probability mass distribution formula as below:

which becomes

n = NP = number of participants
k = f = the number that are female
p = percentage of f in population being sampled

I have calculated Pr (F=1) for KP = 26. Assuming for the moment that p = 0.5 (i.e., that the population to be sampled is 50:50 male vs female) then Pr (F=1) = 3.8743E-07. This is highly unlikely by chance alone. However the assumption of p = 0.5 is certainly off in some fields. I therefore calculated P (F=1) for different frequencies of F in the population (i.e., what is the expected ratio of females to sample from).

Thus for a meeting with NP = 26, only when the frequency of F is ~0.16 does P (F=1) exceed 0.05. So a question is then, what should we use for p for this meeting? An informal survey (John Hogenesch, posted to Facebook at https://www.facebook.com/jonathaneisen/posts/10151208978630767?comment_id=24634832&offset=0&total_comments=15 ) suggests that in qBio the percentage is about 20%. However that may not be an ideal estimate since this meeting is specifically about synthetic biology, I do not have a any estimate of p for this field. However, examination of key meetings in the field (e.g., see http://syntheticbiology.org/Conferences.html for a list) reveals a percentage of perhaps a bit higher. For example at SB5 the ratio was about 35%. I conclude that it is likely that p > 20% in Synthetic Biology. Given that for p = 0.2 the Pr (F=1) < 0.05 I therefore conclude that the null hypothesis (that having one female out of 26 key participants) can be rejected - and that this meeting has a biased ratio of males: females.

UPDATE 2: Here is the full email I received, just for the record

ABSTRACT SUBMISSION DEADLINE 09/15/12http://w-qbio.org/abstracts.html

Quantitative Biology on the Hawaiian Islands
February 18-21, 2013http://w-qbio.org/

The Winter q-bio meeting brings together scientists and engineers who are interested in all areas of q-bio. Each year, the meeting will rotate on the Hawaiian Islands with a different thematic focus within q-bio. The focus for the meeting this year will be Synthetic Biology, with about half of the invited speakers chosen as renowned experts in this area.

SPONSORED BY:UC San Diego's Division of Biological Sciences and BioCircuits Institute
San Diego Center for Systems Biology
University of Hawaii
Office of Naval Research

Jim Collins, Boston University
Johan Elf, Uppsala University
Michael Elowitz, California Institute of Technology
Timothy Elston, UNC Chapel Hill School of Medicine
James E. Ferrell, Stanford University
Martin Fussenegger, ETH Zurich
Leon Glass, McGill University
Terry Hwa, University of California, San Diego
Roy Kishony, Harvard Medical School
Galit Lahav, Harvard University
Andre Levchenko, Johns Hopkins University
Wendell Lim, University of California, San Francisco
Andy Oates, The Max Planck Institute, Dresden
Bernhard Palsson, University of California, San Diego
Gurol Suel, UT Southwestern Medical Center
Chao Tang, Peking University
John Tyson, Virginia Tech
Craig Venter, The J. Craig Venter Institute
Chris Voigt, Massachusetts Institute of Technology
Ned S. Wingreen, Princeton University

Bill Ditto, University of Hawaii
Jeff Hasty, UC San Diego
Bill Hlavacek, University of New Mexico
Alex Hoffmann, UC San Diego
Brian Munsky, New Mexico Consortium
Lev Tsimring, UC San Diego

Registration fee covers conference venue, opening reception, banquet, coffee & snacks.

EARLY BIRD ($450.00) REGISTRATION DEADLINE: December 1, 2012

REGISTER NOW: http://w-qbio.org/abstracts.html

HOTEL: A block of rooms have been reserved for registered conference participants available for a negotiated rate of $199 per night at the Hilton Hawaiian Village in Waikiki. The rooms are available on first come first serve basis and will be available soon, so book early!

CONTRIBUTED TALKS: If you wish to present your work at the conference, either as an oral talk or a poster, you must submit an abstract through the conference website by the September 15th deadline. Abstract guidelines and submission information at:http://w-qbio.org/guidelines.pdf

ABSTRACT DEADLINE: September 15, 2012
Accepted abstracts will be announced October 31, 2012.

We encourage you to forward this message to any colleagues that may be interested in taking part in this exciting event.

Questions should be emailed to: coordinator@w-qbio.org

UPDATE 4:  (9/18/12)

Plus some links that may be of relevance

UPDATE 6: 9/23/12

Some more links on the recent PNAS paper on gender bias and evaluating scientists

UPDATE 7:  9/23/12

Interesting article on gender and invitations to write major reviews

UPDATE 8: More follow up to the Gender Bias study from PNAS 9/26

UPDATE 9: Other posts on gender bias of interest

UPDATE 10: 11/21/13

Just got this in my email.  Kudos to the people behind qBio for adding more women to their planning committee and adding a many women to the speaker list.


UPDATE:  In response to participant interest, the submission deadline has been extended to December 2, 2013.  This year 15 contributed talks will be selected from the submitted abstracts to be presented with the invited talks during the plenary sessions.  Contributed talks will also be selected for parallel breakout sessions which commence in the late afternoon.

Quantitative Biology on the Hawaiian Islands
February 17-20, 2014

The Winter q-bio meeting brings together scientists and engineers who are interested in all areas of q-bio. The venue for 2014 is the Hilton Waikoloa Village, which is located on the Kohala Coast of Hawaii’s Big Island. The resort lets you experience breathtaking tropical gardens, abundant wildlife, award-winning dining, world-class shopping, art and culture, and an array of activities. The Island of Hawaii is the youngest and biggest in the Hawaiian chain, providing a vast canvas of environments to discover--home of one of the world’s most active volcanoes (Kilauea), the most massive mountain in the world (Maunaloa), and the largest park in the state (Hawaii Volcanoes National Park).

UC San Diego BioCircuits Institute and the San Diego Center for Systems Biology
The University of Hawaii at Manoa
UC San Diego Divisions of Biological Sciences and Engineering
The Office of Naval Research

Naama Barkai, The Weizmann Institute of Science
Sangeeta Bhatia Massachusetts Institute of Technology
Hana El-Samad, University of California, San Francisco
Zev Gartner, University of California, San Francisco
Taekjip Ha, University of Illinois
Shigeru Kondo, Osaka University
Arthur Lander, University of California, Irvine
Andrew Murray, Harvard University
Steve Quake, Stanford University
Petra Schwille, Max Planck Institute
Christina Smolke, Stanford University
Aleksandra Walczak, Laboratoire de Physique Théorique

Kevin Bennett, University of Hawaii at Manoa
William Ditto, University of Hawaii at Manoa
Hana El-Samad, University of California, San Francisco
Jeff Hasty, University of California, San Diego
Alexander Hoffmann, University of California, San Diego
Galit Lahav, Harvard University
Eva-Maria Schoetz-Collins, University of California, San Diego
Chao Tang, Peking University
Lev Tsimring, University of California, San Diego

Registration fee covers conference venue, registration reception, banquet, coffee & snacks.

EARLY BIRD REGISTRATION ($500/$425 Student) DEADLINE: December 20, 2013
REGULAR REGISTRATION ($600/$525 Student) DEADLINE: January 31, 2014
LATE REGISTRATION ($675/$600 Student) After January 31, 2014

REGISTER NOW: http://w-qbio.org/

HOTEL:  A block of rooms has been reserved for registered conference participants at a negotiated rate of $199 per night at the Hilton Waikoloa Village. The rooms will be available soon on a first-come, first-served basis, so book early!

CONTRIBUTED TALKS:  If you wish to present your work at the conference, either as an oral talk or a poster, you must submit an abstract through the conference website by the November 5th deadline. Abstract guidelines and submission information at: http://w-qbio.org/abstracts/

ABSTRACT DEADLINE: EXTENDED UNTIL MONDAY, December 2, 2013 (Extended due to large volume of interest!)
Accepted abstracts will be announced by December 6, 2012.  You may submit your abstract now and if accepted, still register by the early bird registration deadline of December 20, 2013.
Abstract guidelines and submission information at: http://w-qbio.org/abstracts/

We encourage you to forward this message to any colleagues that may be interested in taking part in this exciting event.

Questions should be emailed to: coordinator@w-qbio.org


  1. I find it quite incredible that they don't have more women even just by pseudorandom sampling of the field, in which there are so many women. As an exercise, it took me all of 15 minutes to work out of a matching list of 24 women (and one man) that would make a great meeting.

    1. Iddo: Your matching list of 24 women and one man is interesting, I suppose, but does not sum to 26. This is a post about lists with 26 names, and not about lists with any number other than 26 names. I'm with Jonathan on this one and think 26 is a much more interesting number. As he points out, one of the factors of 26 is 13, which also happens to be my daughter's age. You might consider limiting yourself to lists that contain 26 items in the future!

    2. Both the sponsors and the organizers fell down here. I organized my first workshop a few years back; it was part of a larger series. I had a (white male) co-organizer and I'm rather ashamed to say that our first draft of the speaker list was not nearly diverse enough. Fortunately the (experienced) senior organizers of the workshop series were alert, and had received explicit instructions to foster diversity from their funders, and so we were fortunate enough to be given a second chance to fix our mistake (which we did; much to the improvement of the workshop, I might add). One take-home lesson for me (apart from feeling a bit stupid for not having done the right thing from the get-go) was that one cannot just assume that diversity will happen organically and naturally. Privilege is too engrained, even in people who think of themselves as fair, and people tend to forget noble aims when they are under pressure. You need multiple, explicit checkpoints to make sure that diversity and inclusivity are maintained. These checks need to occur not just at the level of the meeting organizers, but also their senior mentors, and also at the level of the funding body (or bodies) that sponsor the meeting. Bottom line: you're entirely correct to criticize the hosts/organizers of this meeting, and I'm glad that you include the higher-up funders in the critique as well.

    3. Incidentally, John: one way that we could ensure future lists contained exactly 26 items would be to stipulate that all list items can only be indexed alphabetically. This will no doubt require a few example-setting denunciations of recidivists who continue to use Roman or Arabic numerals in their lists, but one can hope that only a few such people will need to be drummed out of academia before everyone else gets the message.

  2. Not that it changes the conclusions, but the probability should be cumulative. So P(F=<1) not P(F=1).

    1. good point .. it will not really change the numbers much but I will update

    2. the values for the plot if I did P (F<1) would be

      p P (F=1) P (F<1)
      0.5 3.8743E-07 4.02331E-07
      0.45 3.77792E-06 3.77889E-06
      0.4 2.95675E-05 2.95675E-05
      0.35 0.000191371 0.000191371
      0.3 0.001046034 0.001046034
      0.25 0.004891532 0.004891532
      0.2 0.019645045 0.019645045
      0.15 0.067071458 0.067071458
      0.1 0.186653477 0.186653477
      0.05 0.360606445 0.360606445

    3. OK - apparently my formatting skills in comments suck ... the data does not change much

      for p of

      the previous values for p(F)=1 were
      p (F=1)

      and now they are
      p (F<1)

    4. The good thing about using a cumulative function here is that it could account for speakers that are only partially female. Personally I like to think of myself as at least 5% female (which might be why I was never invited to this meeting). However, if we discount the possibility of partially female speakers and partial speakers and treat each human as a discrete male or female individual then Jonathan's application of the binomial probability mass function seems appropriate.

    5. appropriate assuming you want to know the probability of exactly one female...else maundering's suggestion is helpful since you might be interested in the chance of a speaker list containing 0 or 1 females, which would increase the p-values in Jonathan's analysis ever so slightly. (need to think twice before late-night posting!)

    6. yeah yeah - I think we do want to include the possibility of 0 females ... so I posted the new numbers above ..

  3. This comment has been removed by the author.

  4. I would be happy to collaborate.

    (this is an experiment)

  5. Data for validation:
    A similar lack of female speakers happened earlier in the year with the original speakers announced for "Design, optimization and control in systems and synthetic biology" in Paris.

    This was discussed on the public synthetic biology announcement mailing list

    For those thinking about how to make excuses. I recommend reading the whole thread:

    Looks like the result was that Professor Del Vecchio was invited. See final lineup.

    I'm guessing, but these independent events seem suspiciously related ... :) Thoughts or new theories on the 1 XX speaker?

  6. Ian, I like your insights on diversity. If you decide to leave the field you might do well in non-profit management. I am taking the liberty of quoting you (initially out of context)on Facebook.

  7. The real reason for lack of diversity? Its because all the women who might have attended the Q-bio conference are, in fact, tied up at *another* meeting: http://storify.com/techladymafia/boston-api-jam-s-marketing-problem

    Every time I read that it makes me even more angry...

    1. Holly - link does not come through correctly ...

  8. Brother Jonathan. One does not want to quibble about shades of improbability here. And I'm astonished that Ian did not point this out. But you use the wrong statistics. There is not an infinite population of scientists from which the speakers at a conference should be drawn, and once one is chosen, they can not be chosen again. Thus you should be using the hypergeometric distribution. The probability one calculates is dependent on how big we imagine the population to be. Let's say there are around 2000 people who are in roughly the right field and have the right level of prominence to be selected. If half of them were women. The probability of choosing 0 or 1 in a sample of 26 is ~3.5E-07. Now, of course, that pool is not 50% women (in part because of people being dickheads like these conference organizers). If the pool is closer to maybe 30% women, which seems about right, the probability of 0 or 1 is ~0.001. Since there are lots of stupid conferences every year, this means that we should actually expect one or so every to year to hit this kind of lowpoint if they just chose speakers at random.

    So the organizers could argue that they were just the victims of bad luck in a random selection process. Though I don't believe that for a second. Overt and more subtle sexism is rampant in the sciences. The real question is, no matter what process got you there, how do you look at a list of speakers like this and not say "Oh fuck. We really screwed this up."? Imagine they had randomly selected speakers for a conference and they were all named "Brown". Or they were all from Boston. Or all worked on worms. They would surely notice and do something to fix it.

    Most conferences I have helped organize have consciously tried to have a panel of speakers that are diverse in research topics, geography, stage of career. Indeed it seems that they did all of this when selecting speakers for this conference. That the organizers neither strived to do that with gender, nor noticed when they fucked it up, is a massive condemnation not just of them, but of the extent to which sexism is accepted as a matter-of-fact, not even noticed, part of scientific life today.

    1. Umm ... Michael .. the numbers I have (see table and/or figure) come to about the same as what you report in your calculation

      P (F=1) for p= 0.3 = 0.001046034

      So seems like the hypergeometric distribution with your assumption of 2000 people gives about the same numbers even when p = 0.3.

    2. As for whether P-0.001 should be considered bad luck or just bad - sure there may be 100s or even 1000s of conferences to sample from and thus we expect some to have even this messed up a ratio. So perhaps this could be random flux ... but of course as you point out ... this is not rolling dice ... this is a meeting .. and if your random draw drew such a biased set one should recalibrate if one has any sense ...

    3. Don't worry. I am aware that the hypergeometric and binomial distributions converge for large N. Just taking the opportunity to remind people which to use.

      I honestly can not comprehend how people could not see their own speaker list and be completely horrified.

    4. All corrections appreciated --- I posted my abstract in to be open about it and I hope that if I am selected to present it at the meeting I can make it much better through feedback from the community.

    5. Michael, I am covered in shame for not denouncing the binomial/hypergeometric conflation and thereby implicitly encouraging the horrendous bias towards binomial distributions that arises as a result of "binomial privilege"

    6. Are you excluding intersexual/hermaphrodites from the list of possible genders or are you assigning them as males?

    7. If you read the abstract I noted "I realize there are other gender classes and I regretfully am not including them here" ... so I was not including them

  9. And who could pass up the opportunity to say "hypergeometric distribution" and "total fucking assholes" in the same post

  10. Love your post. This is hilarious, and making quite the rounds among women-in-science email lists...

    Did you really send in that abstract? If yes, I can't wait to hear what the organizers say when they read it. Kudos to you for noticing the horrible skew. I'm not sure many men would have.

  11. lol, they probably only invited that one woman because they didn't realize she had a female name!

    1. That's what I was thinking. It would be interesting to calculate the probability that all females in a list of speakers happen to have a "non-standard" name.

  12. As of 5.35pm PST, the conference programme includes three extra women beyond the list that you posted - Mary Dunlop, Kit Pogliano and Ruth Williams. Maybe they're taking note!

    1. I have posted the full email invitation I received for the record to show that these people were not on the list in the email. Perhaps other people harassed them about this ...

  13. No student registration price? Boo. Maybe I'll try and crash the meeting just to help the ratio a little...

    1. Maybe you can go as me if my abstract gets accepted ...

    2. YES! Though I personally think it would be awesome for your abstract to get accepted. And for you to come out here and present it.

    3. please dress as a woman, if you are selected.

  14. I think you've got a selection bias issue here. The question isn't "what are the odds of a given conference having this sex ratio"; the question is "what are the odds of at least one of the many conferences in your inbox having this ratio". Which is likely to be quite a bit higher.

    For example, if you receive invitations to 20 conferences a year, you'd expect one of them to have a 95%ile overabundance of men giving talks.

    IIRC to get the equivalent of a 95%ile confidence interval, you'd need to take a p-value of 0.95^(number of conference invites in your inbox). But I suck at stats, so check with an expert. Note that this conference might not be significant at that revised level.

    Incidentally, if you actually get accepted to give the talk, it might be worth improving your estimate of the frequency of female quant biologists. You could do this in a reproducible fashion by listing all the US universities with a quant biol faculty and then checking how many researchers they have of each gender.

    Shouldn't take too long. If it does, I'd suggest you set up a Google Docs spreadsheet and ask the Internet to help.

    1. Sorry, 0.95^(1/number of conferences)

    2. Let's put the number at 10. That is about the number of conference invitations where I have looked at the male : female ratio in the last year.

    3. Then the correct p-value is about 99.5%. So, assuming the proportion of females in the QB community is 20%, having only one female is not statistically significant. (Having no females would be significant.)

      Your options are:

      1) If you think there's a systemic bias against women in QB conferences, you could check all of them and aggregate the results using something like http://en.wikipedia.org/wiki/Fisher%27s_method

      2) If you think this conference is significantly worse than others, IIRC you could use ANOVA to confirm this. (Although that would then raise the question: are the others engaging in positive discrimination?)

      3) If you think 20% is too low, you could do a proper survey. The proportion would have to be about 25.25% to get statistical significance.

      4) Finally, you could track down a real statistician and get them to point out the gaping holes in my analysis.

      Whichever approach you take, please be wary of just applying lots of statistical tests until one gives the right answer. If you torture the data enough, it will confess to anything, and that's not a good thing...

    4. Thanks ... probably going to leave it as is ...

  15. The speaker list got updated!

    Likely this post made the change happen. Now the question is how will it look next time or for another q-bio type conference?

    Jim Collins, Biomedical Engineering, Boston University
    Mary Dunlop, Engineering and Mathematical Sciences, The University of Vermont
    Johan Elf, Molecular Biotechnology, Uppsala University
    Michael Elowitz, Biology, Bioengineering, Applied Physics, California Institute of Technology
    Timothy Elston, Pharmacology, UNC Chapel Hill School of Medicine
    James E. Ferrell, Biochemistry, Chemical and Systems Biology, Stanford University
    Martin Fussenegger, Biotechnology and Bioengineering, ETH Zurich
    Leon Glass, Physiology, McGill University
    Terry Hwa, Physics, University of California, San Diego
    Roy Kishony, Systems Biology, Harvard Medical School
    Galit Lahav, Systems Biology, Harvard University
    Andre Levchenko, Biomedical Engineering, Johns Hopkins University
    Wendell Lim, Cellular and Molecular Pharmacology, Biochemistry and Biophysics, University of California, San Francisco
    Andrew Oates, Cell Biology and Genetics, The Max Planck Institute, Dresden
    Mariko Okada, Laboratory for Cellular Systems Modeling, RIKEN RCAI
    Bernhard Palsson, Bioengineering, University of California, San Diego
    Kit Pogliano, Biological Sciences, University of California, San Diego
    Gurol Suel, Pharmacology, UT Southwestern Medical Center
    Chao Tang, Theoretical Biology, Peking University
    John Tyson, Biological Sciences, Virginia Tech
    J. Craig Venter, The J. Craig Venter Institute
    Chris Voigt, Bioengineering, Massachusetts Institute of Technology
    Ruth Williams, Mathematics, University of California, San Diego
    Ned S. Wingreen, Molecular Biology, Princeton University

    1. Still not good enough I note ..

    2. fair point. My comment is meant to illustrate your actions made an impact. There is hope and all progress should be celebrated. :)

  16. yeah jonathan! keep up the good work (expected nothing less) ... as I said before never ever doubt an evolutionary geneticist who recites shakespeare to you atop an amphitheater in termessos slainte me amigo!terrific seeing you again, monica

    1. so great seeing you too --- going post my retroblog from Turkey soon ... be prepared ... and next time you are out let me know

  17. Mr. Eisen,

    Why are you bashing the organizers just based on the list? One could have chosen a multitude of diversity criteria to judge the list (ethnicity, race, religion, nationality). They probably chose speakers based on the merit and relevance of their work to the subject matter (as one would expect). This does not imply any ill-intention or mistake on part of the organizers. Any reasonable minded person would not make baseless judgements simply based on a list of invited speakers.

    1. I think without a doubt a speaker list and organizer list put together with 1/26 being female - in this field - is clearly amiss. I think it very clearly means there is some selection bias going on. I think it is very clear that they did not get a robust sampling of the field in terms of merit I do not claim to know where that bias occurred. It could be at the invitation step.. It could be at the acceptance step. I don't know. But if anyone on biology organized a meeting and somehow ended up with 1/26 female speakers they should ask themselves "What just happened?" and try to figure out why and try to correct it.

    2. Thanks for your reply, but again, I have to point out that although you "think it very clearly means there is some selection bias going on", this is still a pure, unproved conjecture. Secondly, this is not just plain "biology", the name of the conference is "quantitative biology". Perhaps there aren't many female biologists working on mathematical models in biology (I don't know). Even if there are, there is no clear indication that there is selection bias (perhaps the work of the selected speakers was more relevant to quantitative study, or the invited female speakers didn't accept). I am surprised the other commentators on this blog simply followed along with the bashing.
      Also, why should the organizers go back and "fix" the ratio? Like I said earlier, you could chose multitude of diversity criteria to judge the list (ethnicity, race, religion, nationality, sex), but why should any of this be a factor for choosing speakers based on merit and relevance to the field (which should be the criteria)?
      Lastly, did you even bother to contact the organizers directly first (before blogging this and insinuating selection bias) if you felt something was amiss? Perhaps that would have been a better recourse.
      Thanks and good night.

    3. The other commentators did not just follow along but I think they all agree that 1/26 in "synthetic biology" (which is the actual theme of the meeting - not quantitative bio which is the theme of the series) is not even remotely representative in terms of gender. In fact, there are so many excellent women working in synthetic biology that I assert again, something is amiss here. In other words, I dispute your notion that the speakers were in fact selected solely on the basis of merit and relevance.

      As for why I did not write them. I did not feel I had to. I felt like this was something worth sharing with the world. I stand by that decision.

    4. Well, I am simply saying that they are innocent until proven guilty (i.e the speakers were chosen on relevant criteria). Since you are claiming otherwise, the onus of proof is upon you. If all you have to say is " the ratio doesn't look right" then you have only provided us with conjecture, as there are a plethora of reasons why the ratio could be as it is. For example, looking at the profiles of the invited speakers, I have found that many of them have a strong mathematical background and are not necessarily "synthetic biologists" (they could just be targeting a specific group with relevance to syn bio).
      It looks like we aren't going to agree on this, but I wanted to point out to the readers of this blog that it is important critically judge one-sided claims, especially considering we don't have the other side of the story. Isn't such skepticism supposed to be a basic part of doing science? Shouldn't this apply to far-fetched claims like the one made here? Man, I hope nobody here ever serves on a jury!

    5. https://twitter.com/MalePrivilege

  18. You may or may not care that you typo'd "Probability" as "Proability" on the graph

    1. all corrections/comments welcome ... will fix when I have a chance

  19. Dear Jonathan,

    Before I post my comment let me identify myself clearly. I am a physicist/physical chemist turned to quantitative biology problems. I am a faculty member in the Department of Chemistry and Biochemistry at UCSD. I was instrumental in the creation of the BioCircuits Institute as an ORU. I am female. And if you go by cultural background, I am hispanic. So I know all about diversity. Now in response to your first posting.

    The meeting is first and foremost a Quantitative Biology meeting, with Synthetic Biology as a particular focus area (subset) this year. I'm sure you appreciate the difference, and that you also realize the importance of differentiating when determining the input for your model. Being a woman in the sciences, I can tell you that the pool is sadly still quite small, especially considering the caliber of speakers that have been recruited to this meeting.

    I also find your use of statistical analysis to imply a purposeful exclusion of women on our part disingenuous. In order to make this point lucid, I ran your analysis on the ratio of women scientists in the UC Davis Genome Center (of which you are a faculty member). According to the website, the Center has 3 women out of a total of 27 faculty and research staff. I obtained demographic data on the percentage of female graduate students at UC Davis in the field of genetics -- roughly 65%, according to http://graduate-school.phds.org/university/ucdavis/program/diversity/genetics/607 (with similar numbers in other graduate programs across the country). Given the percentage of females in the field, I find that your formula gives a probability of ~10e-8 to ~10e-9 (depending on the assumed size of the pool) of accidentally having a 3/27 percentage of women. Similarly biased statistics can be gathered from other programs and departments you are associated with, such as the Department of Evolution and Ecology and the Department of Medical Microbiology at UC Davis. Following your analysis and logic, the conclusion would be that there is a very strong probability that you and your colleagues are intentionally excluding women from your faculty. I highly doubt this to be the case in a sister institution such as UC Davis! It would be collegial of you to extend us the same courtesy.

    Katja Lindenberg

    1. Katja

      Thanks for your comments. A few notes

      1. I never said that women were being purposefully excluded. I said there appeared to be a biased sampling of women from the field. I still think that is the case. There are many possible explanations for this including that women were invited but said no as I have written about in many other posts about women at science conferences. Whether or not this was intentional is not something I can determine. Certainly I could have been more polite in my comments about the meeting but I stand by the politer parts of my comments - I think that the organizers of this meeting could have and probably should have done a better job at representing women at their meeting.

      I do not support the "caliber of speakers" justification for this ratio at this meeting. There are a large number of incredibly strong women in synthetic biology who could have been invited and in that first email message could have been there if the organizers had worked at it. Are you suggesting that people should ignore diversity completely associated with planning conferences?

      2. I agree with your assessment of a bad ratio of women at many departments at UC Davis. I think without a doubt it needs to be worked on. The Genome Center certainly needs significant work - though since it is not a department it does not have full freedom in hiring decisions. In fact, I think that the diversity in general of many departments at UC Davis needs to be improved. I have tried in the last few years to do my part in achieving this at UC Davis when I have some influence. For example (not that I was responsible for this but I am happy about it) in the Evolution and Ecology department we have made two new hires recently -both women - both brilliant. The ratio in the department is still not ideal nor is the representation of full diversity of people, but we are working on it. I think you are in fact right to point these things out.

      3. When my friend and colleague in the Evolution and Ecology Department Cathy Toft passed away from cancer last year I vowed to work to improve the diversity of women at science meetings because I talked to her extensively about this problem.

      4. As I suggested above, I accept that I may have been too snarky in some of my comments here and the style of my criticism. But I stand by the general point - this meeting had a ratio of males to females that is unsound.

    2. And I note the Medical Microbiology department is actually doing OK in this area - on the web site where you may have looked for the department 17 males, 9 females, and two of the men are not actually faculty in the department. But that may be a bit of quibbling on details. I think there is no doubt that there is a need to work on better inclusion of women and minorities in the STEM areas - in faculty - at conferences - on editorial boards - in major academies - and more.

    3. Jonathan,
      Thanks for your efforts to promote women in the sciences. An area where you could have more influence is in the ratio of males to females on your own Advisory board for PLOS Biology. There appears to be only 3 women out of 25. Not too much better than the Q Bio conference.

      PLOS Biology Advisory Board:
      James Ashe
      Nick H. Barton
      Peter B. Becker
      Margaret A. Goodell
      Douglas R. Green
      William A. Harris
      Daniel Herschlag
      Fred Hughson
      Laurence D. Hurst
      Andre Levchenko
      Michael Lichten
      Michel Loreau
      Georgina M. Mace
      Philippa Marrack
      Tom Misteli
      Eric Nestler
      David Pellman
      David Penny
      Gregory Petsko
      Hidde L. Ploegh
      David S. Schneider
      Ben Sheldon
      Antonio Vidal-Puig
      Matt Waldor
      Peter Walter

      - Elizabeth Thomas
      Associate Professor
      Scripps Research Institute

    4. Thanks Elizabeth. I was not involved in putting together that Advisory Board - it was put together but the professional staff at PLoS Biology (PLoS Bio is run by the professional editors much like Nature or Science. I have an incredibly nebulous role there (see my post about this here ). As my first task at the chair of the Advisory Board I am going to work to have better representation of females - if I have no influence in who is selected (which might be the case) - I will resign.

      PS - Thanks for pointing this out - I confess - I did not notice when I got given a new role as Chair of the Advisory Board.

    5. OK - am working on it ... have sent an email to all of the professional staff saying I want this to be my first task as chair of this Advisory Board. Thanks again for pointing this out ... painful to see ... but clearly needs work.

  20. I wish this were an isolated incident (a radically skewed male/female ratio). Reminds me of an answer I posted on Quora a few years ago, http://www.quora.com/Outliers-2008-Book/How-do-you-know-if-youre-an-outlier. 17 men listed, 0 women.

    Then there's the newly updated Sequoia website, vaunting Silicon Valley founders, http://www.sequoiacap.com/. If I counted right as all those male faces whizzed by, 46 men, 1 woman. Not surprising, given that the firm's partnership has zero women, http://www.sequoiacap.com/us/team/venture, http://www.sequoiacap.com/us/team/growth.

    Don't know what's worse, tech or sciences.

  21. Besides the use of frequentist significance testing (gross), your own data suggest that this event is a fluke, rather than the other way around. You say that most such conferences have ratios quite unlike this one. So they probably reflect more accurately on the probability distribution of conference sex ratios in its entirety. Whereby this event cannot be taken as an instance of systemic bias, at least not in these sorts of events. When you have a probability *distribution*, extreme events, including those that fall in that 5% zone can certainly happen. (Just look at volatility in the stock market, though this admittedly comes from a more uneven distribution than your binomial distribution.)

  22. And re: the pitfalls of frequentism ... I'm not sure ~16% vs. > 20% (but presumably not by too much?) is *practically* significant.

  23. I received an email from Ilya Nemenman:

    Dear Jonathan:

    One of my friends has forwarded me your posts about the gender imbalance in w-qbio conference.

    I think the goal is certainly noble, but I also think a simple binomial model is not capturing the essence of the problem that conference organizers often encounter. I co-organize the original q-bio conference, and we often run into gender balance issues as well there, even with multiple females among the organizers.

    However, this is not for the lack of trying. In fact, we typically invite a lot of female researchers. But they accept our invitations with a much lower rate. I don't know why for sure, but I think that, at least, two reasons are important. First, because there are so much fewer senior women researchers in our branch of science, and because most conferences try to diversify their programs, the female researchers get disproportionately overcommitted. Second, in many cases the unequal child-caring roles between males and females are also a determining factor. At the original q-bio, we try to alleviate the latter by, for example, providing family accommodations, but the success in diversifying has been limited.

    I don't know much about the w-qbio organization process (though I am a good friend of some of the organizers and speakers), and don't want to defend or attack them. However, I wonder if they have run into the same issue. If you add to your model that, among senior faculty, the ratio is probably 90/10 (not 80/20), and that the response bias is between 2/1 and 3/1 (I am guesstimating from my own experience) -- then the w-qbio starts looking much better.



    P.S. Please feel free to publicly share this message on your blog if you deem this useful.

    Ilya Nemenman
    Associate Professor of Physics and Biology
    Emory University
    Atlanta, GA 30322

    1. I wrote this in response


      Thanks for the email.

      I completely understand that women may disproportionally say no to invitations. However, I have organized many many conferences and we anticipate that and try to figure out ways to correct that (such as in ways you suggest).

      In addition, I find the concept that conferences should be about inviting only senior people to be shortsighted. There are many factors that I believe should be involved in selecting speakers for conferences. Only one is to have well established names give talks, hopefully good ones. Another is to give people at different career stages exposure and practice. By expanding out of just senior scientists, one frequently finds a more diverse pool of people from which to sample and this not only can increase the diversity of speakers it can help provide speakers from other groups (e.g., women in some fields) the exposure they need to help them become well sought after senior scientists. Another factor to consider is to even more junior people in the field examples of researchers/speakers at different career stages with different backgrounds (or different genders). I could go on and on. I find it sad that people would use the excuse that "the senior people in our field are not diverse, therefore we can't have any diversity among speakers."

      Thanks again for the comments.


    2. Ilya wrote back

      Dear Jonathan:

      Thanks for the response. I completely agree -- and this is precisely the route we've taken with q-bio, for example. Over years, we have dropped the number of senior invited speakers from 30 to 15 at q-bio, and the junior part of the program is a lot more diverse (and, frankly, a lot more fun, with usually better talks). BTW, coincidentally with your blog post or not, w-qbio seems to have about 5-6 females now on the list of speakers.



  24. I asked around and compiled a list of women related to q-bio. Wanted to share in case useful for conference organizers in general.

    This is by no means comprehensive; apologies to the many people not listed. Perhaps others can add to it.

    Nancy Allbritton
    Frances Arnold
    Annelise Barron
    Bonnie Berger
    Pamela Bjorkman
    Andrea Brand
    Michelle Chang
    Mary Claire King
    Jennifer Cochran
    Virginia Cornish
    Domitilla Del Vecchio
    Terry Gaasterland
    Guri Giaever
    Xinshu Grace Xiao
    Hopi Hoekstra
    Jin Hyung Lee
    Rachel Karchin
    Sobin Kim
    Daphne Koller
    Galit Lahav
    Melissa Lambeth Kemp
    Nina Lin
    Jennifer Linderman
    Susan Lindquist
    Elaine Mardis
    Alexa McCray
    Franziska Michor
    Kristan Naegle
    Erin O'Shea
    Carole Ober
    Sarah Otto
    Linda Partridge
    Dana Pe'er
    Katie Pollard
    Kristala Prather
    Astrid Prinz
    Aviv Regev
    Ann Rundell
    Chiara Sabatti
    Pardis Sabeti
    Silvia Salinas Blemker
    Elizabeth Sattely
    Kristen Scott
    Pam Silver
    Christina Smolke
    Barbara Stranger
    Hua Tang
    Jessie Tenenbaum
    Julie Theriot
    Janet Thornton
    Alice Ting
    Claire Tomlin
    Olga Troyanskaya
    Barbara Wold
    Fan Yang

  25. This is Katja Lindenberg again. As I read this continuing blog I feel the arrogance of it all!! A bunch of (white) middle-aged men throwing around what is right and what is wrong about the presence or absence of women at these conferences!! This is how it goes all the time. A bunch of white middle-aged men talking above and around our heads without asking us where we are coming from. How about an invitation to US to tell you why we are not there as plentifully as you might think perfect? Why not ask us why there are relatively fewer women in quantitative sciences? Where are you coming from to have this conversation without asking those affected? A few women have piped in, but not because you have solicited their opinions. This reminds me of our US Congress talking about.... well, I won't go on. You know what I mean. I think a lot of the comments have been just terribly arrogant and condescending.

    1. Actually Katja I have repeatedly solicited input from women on this issue and related topics. And i continue to do so. Are you saying that because I am a man I cannot comment about cases where I think there are gender ratio issues?

    2. Jonathan is an ally and extremely supportive. He has frequently solicited opinions on this subject from women in science without trying to be the "leader" in the discussion. He does ask those affected--which, by the way, doesn't include only women. We're all affected when women are underrepresented, anywhere.

    3. If the men are the ones leaving women out, it seems like a good thing for a man to point this out and make them (men,as a group)aware of the issue. This is also not the kind of thing where it would seem right for anyone that is left off to make a big deal out of it. If a woman in this field feels slighted by being left off, she would be inviting a whole lot of scrutiny and likely negative criticism for insisting that she should be on the list. I think the average person (men an women) is disinclined to promote themselves in such a way.

    4. Is there a particular person to talk about this; A white young man? A black middle aged man? A white middle aged woman? A black middle aged woman? An Asian, latina, who should talk about it? I thought an expert on a field, regardless of his sex, age, race, sexual or religious orientation can that talk freely about factual information. Katja, your comment was a bit funny though. It made my day.

    5. Personally, I would far rather someone look at this list of speakers and remark on the gender discrepancies in an attempt to raise and address the issue than not, irrespective of that person's colour, race, ethnicity or sex.

  26. I strongly disagree with Katja. As she points out, Dr. Eisen is a white man with every reason to be content with the status quo, and yet he challenges it for the benefit of his field and his colleagues. I think his public efforts should be applauded.

  27. Katja, Jonathan has a voice to which people listen. He's choosing to use this voice to share an observation that he made upon an invitation to speak at a conference. It just can't be the case that you think that only women can speak out about the topic of gender ratio in conference speakers. That, and the tone of your comment, leads me to believe that you must question his personal motives in speaking out on the topic. As a woman who has worked closely with him for many years, I would not argue that he is white, male, or occasionally arrogant, but he is also kind and genuine. He values the contributions of and supports women in science at all levels, from his mentors to his undergrads. You obviously have never met him, or you would never have written those words.

  28. Edwin, thank you, I am glad I made your day! I love making people smile, which I hope you did!

    But Jenna, here is my response to your post.

    FIRST: How nice it is that you feel so strongly and positively about Jonathan! Still, I noticed that in this blog, in all the postings first about the conference and then more broadly about women in these fields, assertions were made that just rubbed me the wrong way. Nowhere were questions asked of or input solicited from those most affected. Maybe I have a different approach, but If I, say, wanted to complain about Latino individuals not being on the list, I would certainly be inclined to ask for the input of Latino individuals and not just simply assert their feelings, say they should be on the list, their life is hard, they are overburdened, they are slighted, they want to be on the list but are not invited. I would want to hear from that group and, if this were my blog (which it is not), I would actively ask!

    SECOND: If my intention was to make things better, I would certainly have communicated with the organizers of the conference before posting such a post! It is true that I have not met Jonathan and that he is probably a totally lovely person. We would probably be great friends if/when we met. I would probably see all the things about him that you say. Still, what he wrote in this particular conversation did not make that clear to me, especially when I looked at the female representation on the Editorial Board and in units that he seems to be connected with at his own institution. He criticized something that was no worse than the units that he could presumably be doing something about. I have been in the UC system for a long time and I know what is possible. That seems not to have happened at UC Davis. I don't appreciate complaining and putting things down just for the sake of creating a row, especially if the complainant has not done better himself. Jonathan wants us to understand that he is not to blame for the composition of his Editorial Board. But if I were to behave and post as he did, I would blame him for that right up front, no questions asked. I complained about this only in response to his post. He did not extend this courtesy to the organizers of the conference.


  29. Katja et al - some comments

    1. You are absolutely right. I could have contacted the meeting organizers to ask them questions about the gender ratio at the meeting. I did not. I note - I routinely send in comments to meeting organizers with such complaints or concerns. I have never gotten a response and have discovered that private conversations about such issues sometimes do not evoke change. So I have started publicly posting about such issues. I agree that contacting the organizers could have / might have been a good thing to do in this case but I still think public postings about gender ratio issues are important.

    2. As for the "Editorial Board" and departments at UC Davis where I am affiliated I generally agree with you that those are also areas that could use improvement (especially the UCD Genome Center as I noted in response to your comment earlier). I don't think it is bad for you to make such comments publicly and I thank you for pointing those things out.

    As I mentioned in a response above, I was not aware of the ratio in the PLoS Biology Advisory Board. I truly had nothing to do with creating that group. I was asked to be the chair of that Advisory Board recently when I raised the issue with the people who run PLoS Biology about not having any actual role there despite my previous lofty title of "Academic Editor in Chief." In reality, professional full time editors run the journal and created the Advisory Board. Whether you believe that or not is up to you but I had no role in selecting the people there. I note - I plan to work to change that ratio and if that is not possible I will step down as Chair. I have as of yet had no insight from PLoS Biology about what "Chair" of the Advisory Board means but if that title can do anything it should have some influence on the gender ratio of the board.

    3. Although I agree with your concerns about gender ratio in the Advisory Board and in some departments at UC Davis,these issues are not equivalent to speakers at conferences. Certainly the ratio in departments and advisory boards is probably more important than for conferences and thus it would be more helpful to the world if that could be fixed. On the other hand, it is much harder to fix and exert ones influence on.

    As you must know from being at UC - one faculty member does not play much of a role in shaping all the hires for a department or research unit. I have tried in dribs and drabs to make improvements but not gotten far. It is not easy. But since it is important, and since you are right that it is not good, I will work harder to see what changes can be made. So I pick on conferences, not departments, because they are so so so much easier to fix. It is usually only a few people who have full authority to pick invited speakers and thus it is much easier to have certain types of balance if that is your goal. So department ratios and conference ratios are just not equivalent things. Again, picking on conference ratios may in fact not be as important as picking on ratios in departments - and I am glad you pointed those out.


  30. As the 'Key People' are actually 2 groups with a strong dependency, a motivated reader might test the follow-on hypothesis: Does the number of females on the conference organizing committee effect the number of confirmed female speakers?

  31. Jonathan,
    Thank you for your nice and candid response. I actually think we would get along and work together really well! Maybe we will meet one of these days. You do sound like someone who is trying to help, albeit not exactly in the way that I try. But our goals are, it seems, the same.

    This is not data-driven, it is a statement I am making purely from my own experience: it is my experience that more diverse organizing committees and search committees lead to more diverse outcomes.


  32. I think Jonathan's instigation of discussions on this topic are very valuable. He has sought input on this topic many times in the past, and his posts are a great forum for discussion.

    I've noticed unbalanced gender ratios (not reflective of the field in question) a few times and when I know the organizers, I always ask about the cause. Of course in some cases, gender balance seemed unimportant to them. However, more than one said that women turned down the offers more often than men. I've certainly limited my travel/activities since becoming a new parent, and I imagine that is a big part of the declined invites among mothers of young kids. I've also heard that women faculty are over-committed because they have extra diversity duties, leading to more declined invites. I've also had organizers say that they (inadvertently) skewed towards inviting older senior scientists, a demographic that is probably more male-dominated. These points were also discussed in a comment above.

    I agree with Jonathan that conference gender ratios are relatively easy to fix -- it may just take a little extra awareness and effort to have a ratio that is reflective of the field's. If women are indeed turning down the invites more than men, then just make extra effort to invite additional women in a second round -- otherwise, even if there is no true discrimination, there is still the appearance of discrimination, which is also no good.

  33. Female graduate student10/08/2012 7:01 AM

    Thank you for this post! Unfortunately, these types of ratios are pretty rampant in this field. So much so, that it was only on the third reminder to register for RECOMB that I even noticed the 15:1 ratio of invited speakers: http://recomb-2012.c2b2.columbia.edu/?q=node/18

  34. Implicit bias is ubiquitous and insidious. But, this reminds me of one of my favorite lines from the African Queen:

    Charlie Allnut: A man takes a drop too much once in a while, it's only human nature.
    Rose Sayer: Nature, Mr. Allnut, is what we are put in this world to rise above.

    blockquote not allowed?

  35. Reading this I remembered that chemistry students always lamented the lack of females in their field as compared with biology. Looked for a conference in chemistry with names of invited speakers being on-line and found this one: http://www.isos2014.de/program/scientific-program/invited-speakers.html

    8 of the 45 invited speakers are clearly female. That's about five times the ratio of the above conference in Hawaii - and that in a field that supposedly suffers from a chronic lack of females.


Most recent post

My Ode to Yolo Bypass

Gave my 1st ever talk about Yolo Bypass and my 1st ever talk about Nature Photography. Here it is ...