I recently read an excellent article in The Psychologist on Bropen Science is Broken Science. It points out that the Open Science and Open Source Software movements, which you would hope might be more diverse and inclusive, are actually less diverse than the general scientific and developer communities*. I want to discuss this and understand it better; I want for us as a community to try and think of ways to root it out.
*The same topic was a major discussion point throughout the launch event for the Essential Open Source Software funding initiative by Chan Zuckerberg Initiative, which I attended in February this year.
Is this real? Yes!
It seems hard to believe, given that these Open Science/Source concepts have been created by communities trying to make things, well, open and free (both in terms of freedom and no/minimal costs). But it’s there, clear as day, in the numbers.
Look at PsychoPy. Within the field of Psychology women dominate the undergraduate population (I’m guessing, but I’d say roughly 80% of students in my undergraduate classes identify as female), which becomes somewhere around 50:50 for postgraduates. By faculty level there are typically more men than women but still quite a lot of women. Within PsychoPy there are surely lots of users that identify as female but when you look at the contributors you find over 125 people, almost entirely male and the vast majority white. The underrepresented groups are there but they are more underrepresented than in the typical set of university professors. To me that’s a surprise.
Does it matter? Yes!
I’m not inclined to discuss this in depth - there are many articles online about the need for greater diversity - but let’s just say I absolutely agree that diversity makes us stronger. We want more diverse opinions about how things should work because software design reflects the designer and if that designer is not representative of the users there’s going to be a disconnect between the users and creators. One of the things I most like about open source software is that the users create it, but if only a small and non-representative subset of users are doing the creating then we might not be ending up with the ideal software package that we all want to exist.
More blatantly, the larger our community of contributors, the better our tools. So, if our pool of contributors is immediately cut in half/quarter, because some demographics feel excluded in some way, then that’s a massive loss to our team!
So why does this happen and can we fix it?
OK, so it’s real and it’s important. Let’s fix this!
…and that’s where I get stuck. I don’t understand it. Of course, I might not be best placed to understand it as one of those that come with the privileged upbringing. Just to be clear I don’t come from a wealthy family - my dad was a clergyman - but I was given a really great (free!) education and the stats are really clear on the advantages of being white and male. So maybe I’m actually part of the problem, but I’m applying myself as best I can to understand the issue so we can fix it.
So here are some thoughts on what could reduce diversity in what is, explicitly, a free and open system. I’m focussing more on the question of the contributors here, not users, where I think the diversity is probably roughly the same as in psychology more generally.
The following are my attempts to brainstorm possible issues. I’m not claiming any of the suggestions are the answer. Just looking for feedback to help us take the right next steps.
1. Is the community welcoming to people of all kinds?
I’ve heard questions about this raised in general although I haven’t personally seen signs of it in the PsychoPy community (again note the caveat above). I think we’ve got a very positive forum where people are very respectful of each other and I’ve seen no evidence of talk that appears in any way exclusive, biased, or even personal. If I’ve missed something, and certainly if there’s something we can do differently, it would be great to hear about it so please get in touch (you can PM me here or email me at jon @ opensciencetools.org) if there’s anything that you think might make some people feel less than comfortable in this space.
2. Is everybody equally allowed to contribute?
One possibility is that people in minorities might not feel like they’re “allowed” to contribute.
To those of us “on the inside” that actually sounds strange. We do actually welcome contributions from everyone, although if it isn’t quite right yet we may need to work with you to get it into shape. But there isn’t, from our point of view, a sense that only certain people can contribute. I’ve never met the vast majority of the PsychoPy contributors. Whether their contribution gets included really will be judged purely on the contribution (is it useful to others, and is it correct?) and therefore it never crossed my mind that people would see it any differently. But that’s me again, not doing a good enough job of imagining how this might seem from the outside. Maybe the absence of diversity indicates (incorrectly) that only some people are welcome. I haven’t spotted the signs, but maybe because I’m not looking for them?
Potential actions? One possibility is that we could run some initiative to raise the awareness of minorities in the community? Maybe we could have one or more “Equality Diversity and Inclusitivity Champion(s)”, volunteer(s) maybe with a little financial bonus, with the aim of raising the visibility of under-represented groups in our community and also to keep pointing out ways we can improve.
3. Is everybody equally willing to be visible?
Maybe a lifetime of avoiding attention, as a minority individual, teaches you not to put your head above the parapet! Maybe you would be more nervous to answer another users’ question on the forum, or to provide a “pull request” to code repository of the software (the code to PsychoPy lives on GitHub and if you want to contribute a change a git pull request is how you request that your contribution gets included).
Let’s be clear that it is scary to contribute to a public project and to have your contribution inspected by the world! The question is whether we can make it less so and improve the environment for everyone.
Potential actions? What we could do here, would be to try and empower people to feel more comfortable making contributions. That’s a good thing, not just for minorities, but for anyone in the community. It’s simply about trying to raise awareness that you have the power to improve your own software! Maybe we could hold some workshops for would-be contributors that want to find out about making their first bug-fix or documentation improvement? Any other ideas?
4. The time to contribute
I don’t think I’ve heard this one expressed elsewhere but it crosses my mind that it might also be a factor.
Creating code that does the job for yourself is one thing. Creating code in a way that others can use it is time-consuming. It means thinking harder about writing code that is general and that others can read, then testing it, and documenting it, and all these things take time. If you’ve come from a disadvantaged background you might be more likely to feel you’ve got to do everything else in order to keep your job and open-source code development might not be top priority for you to keep your job. When someone is more comfortable that their position in life is reasonably secure they can spend time on the “luxuries”, like helping open-source communities.
So, the privileges of being a white male might have got me more quickly to feel sufficiently secure enough in my job that I could dedicate a portion of my time to software development, rather than feeling like I was desperately trying to meet people’s expectations of me. Actually, come to think of it, this might have been a sign of the times too - I think the stress of the academic workplace has increased a great deal in the last 10 years and I’m not sure I would have spent my time on it, if the current pressures were there when I started back in the early 2000s.
Potential actions? I don’t know that I have any ideas to offer if this is the cause. I guess it means fixing the inequalities (or stress levels) in the entire workplace and hoping that they filter through? That could take a while. But am I missing something we can action now? e.g. something to help people do this (or feel they can do this in their current time-constraints?
Do you have other ideas?
I’m keen to hear other opinions on what might be driving the imbalances and, especially, any ideas for how to fix them. If you feel comfortable enough to discuss the issue below that’s great. If you’d prefer to discuss your thought personally with me, that’s fine too.
This is something I’m keen for us all to solve, I just don’t feel at the moment like I know how.
We’ll send that around soon, and hopefully we can retake the survey in a year or two as well to see if we’ve made any difference.