The following recommendations are made for future useR! conferences, as achievable actions to promote the inclusion of women.
Since 2014, 2⁄6 invited speakers have been women. Although a 50:50 gender split is something to strive for, we are aware that this would be a big jump and the focus should be on sustaining the current level in the medium term. The step up in 2014 has already helped to raise the profile of women in the R community.
We recommend that the program committee is selected to be approximately 50:50 men and women. Women have in fact been well-represented in the past, e.g. useR! 2015 had 4⁄11 women and useR! 2016 has 7⁄13 women. Being on the program committee is a boost to the women selected: recognising their expertise, providing experience for their CV, raising their profile in the R community and potentially reducing the cost of their attendance.
Similarly we recommend that 50% of the sessions are chaired by women, ideally with women making up 50% of the session chairs (often people chair multiple sessions). Since the program committee are often called upon to chair sessions, this should follow to some degree from the previous recommendation. However it is also an opportunity to increase the involvement of women on the edges of the R community, e.g. local women who have not previously had much involvement in useR!, regular attendees who have previously not had a formal role, women working in business/industry who are often not considered for such roles, etc.
Not all useR! conferences have panel discussions, but with the exception of the heR panel (“Why aren’t more women R users?”) at useR! 2014, women have barely been represented. We recommend that any panels should reflect the composition of recent useR!s, i.e. at least 20% should be women.
To demonstrate the commitment of the R community to promoting the inclusion of women, we recommend publishing recent statistics of gender balance on the conference website. A good example is set here http://www.lorneproteins.org/policies/. We plan to collate such statistics on the Task Force webpage (https://forwards.github.io/), so that such tables could easily be produced for future useR!s.
Other Issues to Consider
The following sections outline some other issues that we recommend organisers of useR! consider to encourage the participation of women.
We have not evaluated the proportion of women tutors at useR! conferences over the years, but note that 6⁄18 tutorials are being presented by women at useR! 2016. This is in line with the proportion of invited speakers, so seems a reasonable proportion to aim for. This may be hard to put into practice as tutorials are generally contributed rather than invited, however gender balance might be taken into account when ranking submissions or making a final selection.
A large proportion of abstracts contributed to useR! are accepted in some form and there is no evidence of gender bias in the rejections or the assigned presentation format. Quotas might be considered in particular cases to encourage the participation of women. Lightning talks may be particularly suitable for this as the number of abstracts is often much greater than the number of slots. The following links provide good arguments for using quotas: http://womeninscienceaust.org/2015/11/15/achieving-gender-balance-in-less-than-3-years/; http://curt-rice.com/2014/07/03/2-ways-quotas-for-women-raise-quality/. Scope for using quotas may be limited by the gender balance in submitted abstracts: 50:50 may simply not be possible. Alternative actions to encourage the contribution of women might still be considered, e.g. favouring women presenters when ranking borderline abstracts.
Outreach to Women
Efforts should be made to attract women to the conference. An example from useR! 2016 was to email prominent female R users inviting them to the conference and asking them to encourage their colleagues/students, particularly women, to participate. Another idea might be to provide a template email at the end of the registration form to invite a colleague or friend to useR!, with an encouragement to reach out to potential female participants.
If there are local coding groups for women, it is worth considering whether these groups can be involved in any way. There might be a student society for women in computer science or an R-ladies group, for example.
Funding obviously plays a large part in the ability of women to attend useR!. useR! 2016 is offering scholarships aimed at young researchers (funded by the American Statistical Association) and diversity scholarships aimed at minority groups such as LGBT people or people with disabilities, as well as women. Links to any external funding sources might be added to the conference website.
The family-friendliness of the conference should also be considered:
- Is childcare available on campus/at any of the hotels? Such information would be useful to put on the website.
- Could any sponsorship money be allocated to assistance with childcare?
- Is there any potential for a “family room” suitable for participants to take children where (the main) talks could be streamed?
- Are children welcome in sessions?
We hope that following the recommendations and considering the issues above will help to increase the participation of women in useR! conferences. To assist with this endeavour, conference organisers should share the outcomes of any initiatives to increase diversity. In particular, the gender of participants should be recorded to monitor the gender balance of attendees, invited speakers, presenters of contributed presentations/tutorials (broken down by presentation format), session chairs, and members of the local and program committees.