Reactiflux census 2019

Towards the end of 2019, we ran a user survey with the aim of better understanding just who is using Reactiflux (or, at least, who felt comfortable answering a voluntary survey with demographic questions). The results are in.

The results were actually in a while back, but we got caught up with implementing a number of changes to moderation processes and tools, and ultimately delayed this post by a number of months. Sorry!

The questions

All questions were optional, with the intent that if a question made a respondent uncomfortable, they could omit it without being excluded from the results.

• What is your racial or ethnic identity? (multiple choice)
• How old are you?
• What is your gender identity?
• Do you identify as transgender?
• What is your sexual orientation?
• Are you a person living with a disability?
• Do you identify with any of the following religions? (multiple choice)
• Do you feel well represented within the Reactiflux Community?
• Have you felt marginalized or harassed while in the Reactiflux Community?
• Is there anything you'd like to suggest or tell the Reactiflux staff anonymously? (free answer)

We carefully crafted these questions with the aim of being inclusive, and attempted to follow similar patterns set by other demographic-type surveys. If there's a question here that feels inappropriately phrased, or that you think could be improved, we would love to hear from you however you feel comfortable. You can give us feedback in the #reactiflux-moderation channel, DM one of the admins or mods in Discord, or DM @reactiflux on Twitter, or via our anonymous contact form.

The results

We received 545 responses, with the survey open for about 2 weeks. We announced the survey with an @everyone ping in our #announcements channel, and periodically posted reminders in high traffic channels.

We had hoped to receive more recipients. Based on analytics, we know that we have just shy of 300 unique daily message senders, and just over 1000 weekly. At the time, Discord showed online member count peaking at just over 5000 at the busiest times (bottoming out at 2500 or 3000 at the slowest times). 545 survey responses is nothing to sneeze at, but seems to be a relatively narrow slice of those who use Reactiflux.

Reactiflux is predominently young (78% 18-34), white (57%), and male (89%).

Free response

Bucketed into general type of response, we received:

• 22 suggestions
• 58 with positive feedback
• 9 with neutral feedback
• 8 with negative feedback
• 20 who disliked the survey
• 8 non-sequiturs

The suggestions were about anything from new channels, new rules to enforce, new initiatives the submitter would enjoy, or more specific forms of "I don't like what you're doing, please do X instead".

My personal favorite non-sequiturs:

Vegan cheese really does not compare to the real deal

VALHALLAAAAAA

Those who disliked the survey generally regarded it as either an overreach (along the lines of, "why do you want to know who I'm attracted to") or an encroachment of an ideology they opposed (along the lines of, "identity doesn't matter, let's focus on technology").

Those with negative feedback, broadly, had had a bad experience with another member. We're glad to get word of these incidents, but our options for responding to these reports—weeks or months after the altercation—are limited. We took this as a signal that we need stronger mechanisms to report bad behavior within the server, and have made a number of changes since then.

We also received a small number of responses that expressed their frustrating with what they felt was a clique of Reactiflux insiders, both within the MVP/Mod/Admin group and regulars with no specific role. This is a concern of ours, but one that we feel we have to balance with being friendly: we want to avoid cliques that exclude others, but still allow people to be friends within the context of Reactiflux. Any group that has been bound by a single intent for multiple years—in this case, answering programming questions—will inevitably develop some social ties and places of friction, but we want to keep that circle open.

Where we stand

We're disappointed, but not surprised, to see how male-skewed the population is. While digging into other demographic surveys, though, we found these results are mirrored in other large communities.

• State of JS 2018 showed 94% male, 4.7% female (4.3% declined to answer)
• Stack Overflow's Developer Survey was 92% male, 7.9% female

(State of JS 2019 is now at 91%/6.0% male/female, and Stack Overflow's 2020 survey is at 91%/8.0%)

Government data from 2014 (census.gov, page 10. PDF), however, shows web developers as 65% male, 35% female. This is a significant gap—there are substantially more women in our industry than participate in its communities. The conclusion we at Reactiflux draw from this is, there is a persistent force that excludes women these communities.

Reactiflux finds itself in a unique position: we might be the largest realtime chat community for developers, and as such may be able to have a large impact on the broader industry through the norms we set. We're going to take some cues from other large chat communities (Rust, Python, and others), and would love to see any examples of welcoming communities you know of. Drop us a link to them in #reactiflux-suggestions, and together we can make the internet a little bit better.