Mne-python: RFC: increasing diversity of our contributor pool

[drammock]

As some of you know, part of our funding proposal to CZI was for funds to increase participation of underrepresented groups at our code sprints (and in our contributor pool more generally). Due to the global SARS-CoV-2 pandemic, the fall 2020 code sprint in Seattle is postponed, and it is unclear whether a spring 2021 sprint will be possible / whether people will feel comfortable enough with air travel to make the journey. This issue is for brainstorming ways to recruit and retain a diverse pool of contributors, without relying on (subsidies to attend) in-person code sprints. @britta-wstnr and I have been chatting a bit about this; here is a summary of our ideas so far:

1. Prominent public statements that we welcome all new contributors. Locations might include the main homepage, the GitHub README, the MNE-Analyze email footer, and/or including a slide and verbal statement in any MNE-Python-related talks/trainings done by core devs.
2. Designated "ambassadors" who are willing to work directly with new contributors a little bit more than the usual PR review. Ambassadors should be willing to exchange a few emails or videoconference with new contributors to answer questions, give guidance, even do a little screen sharing or pair programming as appropriate.
3. Designated "mentors" who would commit to a certain number of hours/week for a defined period, and get matched with an "intern" to work one-on-one with during that period. It may or may not be possible to provide stipends for the mentor and/or the intern for this option.

The goals for new contributors are flexible and expected to vary depending on the new contributors' past experiences and current interests, but generally will include some of:

• developing comfort and fluency using git / GitHub
• understanding Python package installation (i.e., different ways to install MNE-Python and its dependencies, upgrading, etc)
• learning best (or "good enough") practices for scientific programming & scientific software development (documentation, testing, deprecation cycles, etc)

Please weigh in here if you have alternatives or additions to suggest, want to express support or lack of support for any ideas already mentioned, or have other related comments to share. We are very much open to other ideas. Also please tag anyone you think should see this but who might not get auto-notified without an @-mention.

[hoechenberger]

tagging @SophieHerbst and @dnacombo
cc @peircej @m-macaskill @opyh

[agramfort]

an idea... have like office hours 1H every week or every two weeks to onboard
people or ask questions in a video chat. What you know people it's always
easier to work together

[britta-wstnr]

Thank you for the summary, @drammock.
I think what we should add to the first point (Prominent public statements) is to also make clear that diversity is a value of the MNE-Python community and that we believe it contributes to our joint goals.
To essentially go beyond "diversity just for the stats" or other negative descriptors that can often be associated with quotas or similar - to show that we believe in the positive effects of a diverse contributor pool and therefore actively work on increasing the representation of underrepresented groups.

[drammock]

we believe in the positive effects of a diverse contributor pool

:100:

[jasmainak]

Just a comment as I was passing by. I think it might be a good time to review what issues are truly "easy", which ones are "good first issues" etc. Something I've found helps new contributors is to give them progressively harder issues. If you accidentally take up an issue that's too hard, it's going to discourage future contributions.

[virvw]

Hi folks,

great to see this happening! :-)

My 3 cents:

• inverse the balance and strongly highlight contributions of the
  minorities (even if in the grand scheme of things they are smaller). By
  default, biases favor so strongly the majority that the regular pat on the
  back for a minority is not effective enough. The bragging needs to be 2 or
  3 times louder for minorities...

• What is effective is to be fact-oriented: count the ratio of selected
  minorities in MEG/EEG/... and compare with your current pool. Then state
  out loud: there is 30% of women in the field, we want 30 % of women
  contributors - who is up for it?; there is 5 % of Chinese users, we want 5%
  of contributors etc... In other words, pro-actively seek and invite them.

• Build a system of mentorship: the youngest arrival is paired with a more
  senior for a few months (perhaps what Daniel referred to as ambassadors).

Cheers,

Virginie van Wassenhove
https://brainthemind.com/ https://brainthemind.com/

On Wed, Sep 9, 2020 at 2:34 AM Mainak Jas [e-mail] wrote:

Just a comment as I was passing by. I think it might be a good time to
review what issues are truly "easy", which ones are "good first issues"
etc. Something I've found helps new contributors is to give them
progressively harder issues. If you accidentally take up an issue that's
too hard, it's going to discourage future contributions.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
https://github.com/mne-tools/mne-python/issues/8221#issuecomment-689225936,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABYMUIUWL3DNHZMPJFAFUADSE3ETDANCNFSM4RAI2WZQ
.

[SophieHerbst]

Great initiative and great suggestions above! Something I would think might help too, is to offer small hands-on technical tutorials for new contributors (~2h with an experienced programmer, possibly virtual but interactive) to help people get up to speed and increase their _confidence_, for instance:

• how to use git / github
• principles for clean coding in python
• introduction to the mne architecture
• logic behind the testing (CIs and stuff)

[britta-wstnr]

@SophieHerbst do you mean something like a regular little hands-on lecture new contributors can sign up for? I think that's a great idea!

[SophieHerbst]

@britta-wstnr yes, something shorter than a full workshop (to be integrated in one's daily schedule more easily), which targets the skills specifically needed to contribute to mne python.

[drammock]

something shorter than a full workshop, which targets the skills specifically needed to contribute to mne python.

I like this idea a lot. Perhaps we could start with the framework of The Carpentries lessons and tailor them to our needs.

[hoechenberger]

I just opened my first issue at the NeuroKit2 repo – and got a friendly automated welcome message from a bot! Interesting idea!

https://github.com/neuropsychology/NeuroKit/issues/392#issuecomment-738694192

[cbrnr]

😄 - except for the meme I like this idea!

[hoechenberger]

😄 - except for the meme I like this idea!

Yes :)
One could then even include a link to our forum and such.

[m-macaskill]

That's great - even though it's automated, getting an immediate friendly response is a nice touch.

Mike

On 4/12/2020, at 23:11, Richard Höchenberger [e-mail] wrote:

😄 - except for the meme I like this idea!

Yes :)
One could then even include a link to our forum and such.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or unsubscribe.

--
Michael R. MacAskill, PhD 66 Stewart St
Research Director, Christchurch 8011
New Zealand Brain Research Institute NEW ZEALAND

Senior Research Fellow, michael.[email protected]
Te Whare Wānanga o Otāgo ki Ōtautahi Ph: [phone]
University of Otago, Christchurch http://www.nzbri.org/macaskill

[hoechenberger]

@peircej recently posted "[on] improving Equality, Diversity and Inclusivity (EDI) in our community"
I believe this is a valuable read:
https://discourse.psychopy.org/t/on-improving-equality-diversity-and-inclusivity-edi-in-our-community/18356