I am part of the organizing committee for a workshop in a STEM field that historically has problems with underrepresentation of women (and other groups, but let me focus on women in this question). I would like to help foster an inclusive environment at the workshop and more generally in my department and am looking for suggestions for how to proceed.

The topic of the workshop is slightly out of my field of expertise and I do not have a long list of qualified speakers (of any gender) that I can offer suggestions from.

I suspect that the senior member of the committee will take the attitude that he is "gender blind" and chooses speakers to invite based only on their quality but that because of implicit biases, the invitee list he draws up will be something like 90% male. I do not think he will be particularly open to a direct conversation about gender and underrepresentation.

I saw this question which was asking whether preferential invitation of female speakers is normal; I am instead asking how to foster diversity (perhaps by preferential invitation or by other means).

  • 18
    Note that an unbiased selection of speakers may also come up with 90 % males – if (roughly) 90 % of potential invitees are male.
    – Wrzlprmft
    Commented Dec 29, 2014 at 9:02
  • 11
    @Wrzlprmft It has been suggested that this is a self-perpetuating situation, ie that part of the reason 90% of the potential invitees are male is that 90% of invitees at previous conferences were male. There may be biased sampling at an earlier point, or rounding errors could always fall one direction, creating unintentional bias over time.
    – Jessica B
    Commented Dec 29, 2014 at 9:12
  • What is the procedure by which the final list of invitees will be determined? You mention that the senior member will "draw up" an invitee list, but is that the final list, or just a "menu" from which the rest of the committee will choose? It is possible to have a blind stage followed by a stage in which gender is considered, meaning that women get no bonus when considering whether they are qualified, but, given that they are qualified, get a bonus in determining whether they are actually invited.
    – BrenBarn
    Commented Dec 29, 2014 at 19:34
  • 1
    @Wrzlprmft let me be more specific. I believe that there are many qualified female researchers that one could put on an invitee list but I fear that for various reasons, such as those offered by Jessica B, the list made by the other organizers will only include a few (if any) of them. That is, I am fairly certain that the shortlisting procedure is biased. Because this workshop is outside my area of expertise, I don't know enough to suggest these overlooked female researchers (or male researchers for that matter) myself.
    – anonymous
    Commented Dec 30, 2014 at 2:16
  • 1
    @BrenBarn I like the part of your suggestion that prevents the appearance of tokenism. I'm afraid in this case that the output of the blind stage will contain disproportionately few female names --- not because there are no good female researchers in this area, but because of the implicit biases of the other members of the organizing committee. Then maybe only the very top tier of female researchers are in the shortlist along with the second and third tier of male researchers. Do you have any suggestions for anything that can be done earlier in the process?
    – anonymous
    Commented Dec 30, 2014 at 2:21

4 Answers 4


Aimed at a different point in the pipeline, you can try to make it more possible for speakers to accept your invitation. One step that can make a huge difference for certain people (in many cases, converting a 100% impossibility into an acceptance) is to provide resources for childcare at or near the conference.

For some ideas about specific steps you can take, Matilde Lalin recently wrote a nice overview of different ways conferences can and do support attendees traveling with children (published on Terry Tao's blog).

One that's not mentioned there is to allow attendees to purchase a hotel room in the conference-reserved block for a nanny. I know that the American Institute of Mathematics does this, and it can make the difference if the only other option would put the nanny somewhere far from the conference.

Depending on the size of your conference, this needn't necessarily mean that you subsidize or even organize the childcare; it can help even just to put attendees in touch with local daycares/nannies/babysitters, or provide other information. However unless your conference is very small, I hope you'll give thought to implementing some of the other steps that Lalin outlines.

Added February 2015:

Greg Martin has written an excellent paper "Addressing the underrepresentation of women in mathematics conferences" that addresses exactly this question, which is simultaneously scholarly and actionable. In particular, in Section 4 (pp 17-21) he gives thirty-eight concrete suggestions that organizers can follow. Not all of them will be applicable in a given situation (including that of the original question here), but many will be.

  • Thank you for these helpful suggestions. This focuses on a part of the process I had not previously considered.
    – anonymous
    Commented Dec 30, 2014 at 2:22
  • This paper looks like a great resource. Thank you very much for following up.
    – anonymous
    Commented Feb 27, 2015 at 0:07
  • 5
    Frustratingly there's a logistical problem here, which is that US government stupidly decided in the early 90s that childcare is not "a necessary expense" and so federal funds (e.g. NSF conference money) is never allowed to reimburse for it. So this requires proactively finding separate funds, perhaps from your university. See more here. Commented Dec 10, 2017 at 19:44

This idea is based purely on my own experience:

As a young researcher, cost is a reasonably significant factor in deciding whether to attend a conference. In some settings the cost can be reduced if you can find someone to share a hotel room with. Finding such a person though can be difficult, particularly if there is no public list of who is attending the conference and your gender is under-represented. I think it might be helpful if the registration form had an option along the lines of 'I would like to be put in touch with people who might be interested in sharing a room' (with appropriate follow-through, of course).

  • 2
    Thank you for this suggestion, which I had not thought of. My situation may be unusual as we anticipate covering housing costs for all participants but I'm sure this answer will be useful to other people with similar concerns.
    – anonymous
    Commented Dec 30, 2014 at 2:25

One helpful thing is to find lists of women in the relevant field to look through to find potential participants to invite to speak. For example, in Number Theory you can look at the list of participants in the Women in Number Theory conferences. Not every field has such easy sources of lists, but there are a lot of them out there, and it can help to find speakers whose research interests you but who you hadn't met before or otherwise didn't come immediately to mind. Similarly, social networks being what they are, having women co-organizers of the conference can also be quite helpful in brainstorming great women speakers who you might not have thought of on your own.


First of all, the suggestion of making sure there is childcare and it is possible for everyone to attend the conference is very valid and relevant.

Second, is there more than one person in the committee who can suggest experts in the field? If you suspect that the most senior person will claim to be "gender blind" while not being so, can you involve more than one person in the selection process? Not only this is good to mitigate everyone's potential biases, but it's good too because more people can have potentially more connections. Hence, the more people you involve in the selection committee, the longer list of experts you can reach, and therefore you can possibly aim for a higher quality of speakers. If only one person makes the selection, this person will likely overly represent their collaborators, coauthors, friends and the people they are more connected too, while forgetting about other experts that they may not know as well or not know at all.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .