Improving diversity

[danielskatz]

Discussion topic: software community diversity Brief description of issue/challenge: software (at least in the US) is developed and maintained by a community of people who do not match the overall diversity of the country - How do we improve this? Lead/moderator: maybe @Dr-G ? Links to resources: https://wiki.numfocus.org/DISC:_Diversity_%26_Inclusion_in_Scientific_Computing

[bangerth]

This is a topic we've really been struggling to understand in deal.II. Women (and presumably minorities -- though these are more difficult to spot by name alone) are quite rare among the people who ask questions on our mailing lists. It may of course be that they are there, just don't ask questions, but when we ran a survey, only about 10% of respondents were women -- in other words, even in anonymous settings, women are quite underrepresented.

At the same time, of course, my whole field of computational science is heavily dominated by men. It is quite possible that the composition of our users, our mailing list participants, our leadership, is actually reflective of our bigger community. If that is the case, then the problem is not with our project, but with the community as a whole.

In the end, this does not make the situation any better. It just means that we as a project are not immediately to blame, just we as a community. I would love to understand these issues better.

[ljhwang]

Perhaps this is better framed as a pipeline problem. Can we understand at what career stages diversity is lost or maybe never achieved?

[bangerth]

I'm sure that it is a pipeline problem, and I'm sure that the pipeline is leaky at every stage. It is striking to me that female applicants for math grad school overwhelmingly go into pure math, leaving very few in applied math. Of course, the international applicant pool is heavily skewed towards males as well (in addition to being skewed towards applied math), so they do not make up for the lack of diversity.

I do not know of a good explanation why female grad student applicants prefer pure math over applied math, other than seeing that the fraction of female professors in applied math is even smaller than it already is in pure math.

[bboscoe]

Talk to me, I can offer you a single data point on this.

On Mon, Apr 9, 2018 at 2:59 AM, Wolfgang Bangerth notifications@github.com wrote:

I'm sure that it is a pipeline problem, and I'm sure that the pipeline is leaky at every stage. It is striking to me that female applicants for math grad school overwhelmingly go into pure math, leaving very few in applied math. Of course, the international applicant pool is heavily skewed towards males as well (in addition to being skewed towards applied math), so they do not make up for the lack of diversity.

I do not know of a good explanation why female grad student applicants prefer pure math over applied math, other than seeing that the fraction of female professors in applied math is even smaller than it already is in pure math.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/si2-urssi/berkeley_workshop/issues/11#issuecomment-379614789, or mute the thread https://github.com/notifications/unsubscribe-auth/ALtLv62dAeOh5dYvXvWDB3h_FYxQGudHks5tms6igaJpZM4S7JlN .

[marylhaley]

At NCAR we are dealing with the lack of diversity in our staff and in the candidates who apply for our open positions. Part of the issue is the current environment may not feel welcoming to women and minorities and hence they simply don’t want to apply for jobs. Companies have to make a real and honest effort to change the culture in the workplace.

NCAR hired a Chief Diversity, Equity, and Inclusion Officer who has tackled this issue head on by offering educational programs tailored around making the NCAR work environment more welcoming for people from all backgrounds. A workplace culture survey was conducted across the organization to identify the problem areas and the results were a real eye opener. We are slowly changing hiring practices, in part to change how jobs descriptions are written so they are more appealing to candidates from diverse backgrounds. I think we are trying to get the job descriptions advertised in a more far-reaching manner.

I do think this is a pipeline issue as well, because at many stages (K-12, college, early career) we may be losing some truly qualified people simply because they didn’t feel welcome in their environment.

[bangerth]

I'd like to talk to all of you about your experiences, in particular what the survey results were. I appreciate you being willing to share your experiences!

[ljhwang]

I never considered how to make a job ad more inclusive. I would love to learn more !

Lorraine Hwang Sent from my iPhone

On Apr 9, 2018, at 7:48 PM, Wolfgang Bangerth notifications@github.com wrote:

I'd like to talk to all of you about your experiences, in particular what the survey results were. I appreciate you being willing to share your experiences!

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[danielskatz]

We’ve tried using http://gender-decoder.katmatfield.com/

On Apr 9, 2018, at 19:53, Lorraine Hwang notifications@github.com<mailto:notifications@github.com> wrote:

I never considered how to make a job ad more inclusive. I would love to learn more !

Lorraine Hwang Sent from my iPhone

On Apr 9, 2018, at 7:48 PM, Wolfgang Bangerth notifications@github.com<mailto:notifications@github.com> wrote:

I'd like to talk to all of you about your experiences, in particular what the survey results were. I appreciate you being willing to share your experiences!

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/si2-urssi/berkeley_workshop/issues/11#issuecomment-379957399, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ACx2NWmuBiiXL1Z-bHlqSVh_N7ydFHYNks5tnB6ngaJpZM4S7JlN.

[marylhaley]

I'm happy to talk with folks, but it's more from a witness viewpoint. I've been attending the workshops and seminars that our DEI officer is organizing, and they are incredibly helpful. Our DEI officer is gently encouraging us to take a harder look at ourselves and our practices, and is the key, I believe, to making sure we shift our culture. I'm not sure it could happen without her.

[Dr-G]

Happy to lead/moderate this. Definitely have plenty to say about it!

[abbycabs]

re: job descriptions -- @ljhwang I use Textio

[11fish]

Leaving job ads open longer improves the diversity of the respondents.

[katyhuff]

Something to think about on this topic is how inclusion and the process of diversification are perhaps more important as goals than metrics like percentages (10% respondents, community members, etc.)

This is all along the lines of Goodhart's Law:

Goodhart's law is an adage named after economist Charles Goodhart, which has been phrased by Marilyn Strathern as: "When a measure becomes a target, it ceases to be a good measure."

Additional thought to expand on this: There are types of diversity that are "invisible" which may be worth discussing and diversity along these axes will be hard to measure but may be improved through a focus on inclusion. (e.g. disability, economic class, nation of origin, undisclosed sexuality, etc.)

[Dr-G]

Breakout Group 1 google doc: https://docs.google.com/document/d/1iPVyJuy-ujz99EGIvLEBfkgH7j7mdtO-ZRe9VfVF5Ag/edit