[OpenStack Foundation] [Diversity] Stats on diversity in OpenStack

APlimpton aplimpton at gmail.com
Tue Jun 30 17:14:52 UTC 2015

Core review teams! Thank you Eoghan :)

I would suggest any polling ask for the names of the people. We can then
match that against how the individuals identified themselves during signup.
We want the self-identified values over ones assigned by someone else.

The information in the company field is a useful data point both internally
and externally facing efforts on our part. It is likely populated with
company names, meetup names (maybe, we'd have to see the data), and
independent identifiers (self, none, student, etc). We are trying to grow
and retain diversity in the OpenStack community. That growth in diversity
comes from those three primary sources: companies supporting their people
joining, meetup groups that nurture and retain people and independents who
join without affiliation to a company or meetup. By looking at the
diversity info of those three sources we can evaluate how and where we
should be implementing goals to increase diversity.

The retention comes from making a welcoming and rewarding environment
inside OpenStack. Combining company field data with the other diversity
data provides some insight on tailoring our goals and programs to motivate
and reward in ways that are meaningful to those community members.

Additionally when we have data on companies, and possibly meetups,
consistently doing well at growing and retaining diverse teams we can
reward them -- award, panel at summit-- and/or ask what is working for
them. They can also be potential allies in workshops in regions helping
make sure material is language and culture appropriate.


On Tue, Jun 30, 2015 at 1:21 AM, Eoghan Glynn <eglynn at redhat.com> wrote:

> > Hello,
> >
> > I think one starting point is a snapshot (as much as we can) of what the
> > diversity is like in OpenStack. We have some stats in the etherpad about
> > the Summits -- is there a official document with those numbers?
> >
> > Summit attendees is one metric, but includes a range of people in
> > non-technical roles and possibly not part of OpenStack.
> >
> > Better metrics would include the active members of the OpenStack
> community
> > -- contributors, leaders and management. One measure of active members
> > would be the list of active technical contributors (ATC) which should
> > capture contributors to documentation and development of the projects. We
> > likely need a way to count those contribute in other ways, but for now
> the
> > ATC is a start.
> >
> > I see that the OpenStack signup form asks about gender and location. When
> > was the gender question added? When was the location question added? My
> > concern is that if they were added recently that few records will have
> the
> > data.
> >
> > 2015 Metrics on Gender and Geography Diversity in OpenStack:
> >
> > 1. How many total active technical contributors (ATC) are there?
> >            a. If I read http://www.stackalytics.com correctly, that
> would
> > be 1316
> > 2. What is the total ATC breakdown by geography and gender?
> > 3. What is the per project ATC breakdown by geography and gender?
> > 4. What is the per company ATC breakdown by geography and gender?
> My recommendation would be to keep this centered on the community
> structures and not include company affiliation (since it's the former
> that we, as a community, have the most direct common leverage over).
> > 5. How many project team leads (PTL) are there?
> > 6. What is the PTL breakdown by geography and gender?
> > 7. How many OpenStack board of directors (BOD) members are there?
> >           a. https://www.openstack.org/foundation/board-of-directors/
> has 21
> > 8. What is the BOD breakdown by geography and gender?
> >           a. https://www.openstack.org/foundation/board-of-directors/
> shows
> > possibly               2 women and 19 men.
> >
> > Did I miss any OpenStack metrics? Who would be able to help get the data
> > for the stats?
> The per-project gender breakdown for core reviewer teams would I think
> be even more useful than the PTL breakdown (given the evolution in the
> PTL role over the last 4/5 cycles).
> I can offer the stats for the two projects I'm most familiar with:
>   ceilometer: 3 women,  8 men
>   nova:       1 woman, 13 men
> My feeling (based on no more than anecdata) is that with the exception of
> horizon-core and openstack-doc-core, the ceilometer-core group is at the
> more balanced end of the spectrum.
> One interesting thing would be to correlate the number of cycles that a
> project participated in the OPW/Outreachy program and the number of female
> cores on the project. (Not necessarily that this participation was driven
> by the female cores, or even directly produced female candidates for core,
> but rather that it reflected a more "welcoming" atmosphere on the project
> in some vague sense). In the case of ceilometer, IIRC we mentored OPW
> interns for havana, juno and kilo.
> Perhaps the easiest way to collect those stats from the nearly 40 official
> projects would be to mail the dev list requesting that all PTLs self-report
> the gender breakdown for their projects.
> Cheers,
> Eoghan
> > Thanks!
> > Amanda
> >
> > _______________________________________________
> > Foundation mailing list
> > Foundation at lists.openstack.org
> > http://lists.openstack.org/cgi-bin/mailman/listinfo/foundation
> >
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/foundation/attachments/20150630/64676de8/attachment-0001.html>

More information about the Foundation mailing list