GSoD p5 Tech Writer: General Report

An Overview of Intentions and Findings

A Brief Introduction

This document aims to provide descriptive documentation of my time as a p5 Tech Writer. It contains conversational language around my approach, findings, intentions, successes and stumbling blocks.

Overview of Deliverables

This document contains links to the following assets. Each edit lives on a separate google doc page:

A copy edit suggestion of several pages on the p5 website:

• Get Started — p5js.org/get-started/
• Community — p5js.org/community/
• Teach — p5js.org/teach/

A copy edit suggestion of several pages on the p5 Github

• Code of Conduct — github.com/…/CODE_OF_CONDUCT
• Contributors — github.com/…/contributor_docs/README

A digital zine that compiles some findings into a shareable format that can help other organizations think through their strategies for creating inclusive language on their platform.

Objectives

1. To complete an visual and textual edit of a subset of the pages on the p5 website with consideration towards the following parameter based on historic, market and ethnographic research:

• Language: beginner friendly, approachable and referencial of other non-technical skills
• Tone: welcoming, outlined on the community page, demonstrative of a diverse batch of users
• Engagement: calls to action that open various spaces to continue learning at multiple intensity level
• Advancement: clarity of instructions for contributing
• Layout & Assets: demonstrative of core values

2. To research strategies that other orgs in creative tech have done to take their code of conduct and distribute those core values across their website. To interview existing users and prospective contributors on the ways in which the code of conduct values are visible throughout the website and Github.

Methodology

In this research, I decided to attempt an ethnographic approach to my research. I’m not trained as an ethnographer but this human, interview and conversationally centered style of research is something that is very interesting to me. I am in the process of investigating a pHD in this area but in the meantime, with the small training that I have (via a workshop or two that I have taken) I largely see ethnography as speaking to a lot of people about their lived experience and trying to document those findings.

A Brief Note on Methodology Discoveries

People will tell you a lot of things that you didn’t ask for — which is actually great albeit difficult to document with consistency. I was hoping to humanize the experience of learning about the language shortcomings [on the p5 website] for beginners by having small conversations instead of a survey. There’s nothing against surveys but I was hoping to pull out some of the emotional stressors that come with being a new coder or contributor that, in my experience, are easier said to a person who opens the opportunity with empathy. In this process of hosting small discussions I learned that it is very challenging to ask questions that result in anecdotal data that is easy to map. As I said, I’m not trained in this field so perhaps this is obvious but it was abundantly clear to me throughout this process.

I also learned that often new learners or contributors are still developing the vocabulary for some of the foundations that are dissuading them from creative coding. I noticed that this sometimes meant that I would offer up technical vocab suggestions to help them describe their roadblocks which certainly seems like a way to sway data but felt like the human thing to do in a conversation. I’m still trying to strike that balance of detailed research questions, with space for clear responses. I’ve included some simple examples below. My suggested term is bolded.

• ie. Me: “Do you mean that the interface makes it hard to navigate github?”
• ie. A new contributor: “I just don’t know where to go to initiate an edit” Me: “do you mean the pull request process? Yes, I know that can definitely be confusing.”

There were many other small moments of clarity along the way but the last major commonality from interview to interview was the relationship between creative coding and identity. Questions about whether this space was right for the community members came up time and time again, particularly in the anecdotes from people of colour (but not as often from people of Asian descent) as well as women. Most, if not all, people in these demographics that did not feel a sense of worry around belonging made reference to a time in which they saw evidence that people of a similar identity were working with p5 and/or creative coding. This is not brand new information. Studies have shown for decades that representation can sway the ways in which lessor resourced communities are invited into tech space. What is noteworthy here is that, even within an org that has taken great effort to be more inclusive this anxiety for new learners outside of the majority is still present.

Though my research did not include a comparative analysis between p5 and, say, a computer science program, with all of the standard social ecosystem issues, I would still say that p5 is doing better than most places! This kind of work that engages inclusivity is not about perfection so much as slow, intentional progress. So the question becomes, based on the outcomes from the interviews conducted, what measures could be made to continue to adjust the language to feel more inclusive to new learners and new contributors, with a focus on women and people of colour? I have added some general recommendations based on my market research and ethnographic research in this document.

Audience

I’d like to take an opportunity to discuss a little more about my audience and how I came to know the people that I interviewed. Audiences can also create biases that I think are good to name upfront. I teach at many universities, all of which have graduate and undergraduate studies in creative technology. I had initially set out in this Tech Writer position to connect with a broader range of people, namely people well outside of my network. An unexpected need to tend to some unwell family members as well as difficulty in scheduling around a very busy new job meant that most of the hours served in this role were done in the wee hours of the night between 10pm and 2am or in the odd daytime hours between my classes and tasks.

Though it wasn’t initially my plan, it turned out that it was much easier to sync up with folks that were on a similar schedule as me. This meant that many of the people in my interviews were people who aim to be a part of the creative tech industry, particularly those who have shown interest in contributing or becoming more fluent in the p5 platform. Many of them are students and academic colleagues or friends of students and colleagues. I kept my closest contacts and the students I had most recently taught out of this process as they know too much about my admiration for p5 and my personal distaste for exclusionary tech spaces — I think it would have shifted their response significantly.

Many of the participants answered an email or Discord call for participation from me. Those calls were shared with their peer groups outside of the universities as well. This was very helpful as it allowed me to lean on my community to self-organize our meetups to some extent. This would have been far less possible with a broader public call. The benefit to having some proximity to the audience that I chatted to was that they felt comfortable enough to tell me about some of the hard truths of their experiences in feeling welcomed or included in the creative technology community, particularly p5. I’ve ensured that all of their comments were kept private if they so requested (most did, I suspect many would like to work for p5 one day) but disclosed that I would be incorporating their feedback into a recommendation for the copy edit on the website and github.

In my calls for participation I asked for some trust from lesser served communities and vocalized an intention for a safe, anonymous place to voice their lived experience. An average of 83% of my conversations were held with people who do not identify as male. An average of 58% of the conversations were held with people who do not identify as white. I think that tech spaces, particularly tech art spaces

Overall I was able to meaningfully meet with over 70 students, colleagues, young adults and tech educators over the course of the six month contract. There is some lost data here from a block of sessions that was stored in a file that went missing in a computer crash. The numbers are slightly higher than depicted but I wanted to document what I know for certain. These sessions consisted of a few different meeting styles, as described below. Participants could opt into whichever they wanted:

Style 1: Casual Conversations, One on One
Stat: 10 conversations with 10 people

This style of conversation wasn’t widely selected over the six month period. Many of the folks who chose this method (as well as style 3) were moderately experienced coders who were interested in providing feedback on the experience of contributing to p5. This feedback was clear and easy to map. The reflections are largely shown in my edit of the github copy.

Style 2: Casual Conversations in Small Groups
Stat: Around 19 conversations with an average of 3 people per session, around 57 people total

This style of communication was the most popular. It was low commitment and the social buffer created a low pressure conversational environment. As I learned how to ask the right questions, I tightened my process from session to session. Most people requested that their identity be kept anonymous. I took hand-written notes in order to try and stay present and conversational. I later tried to compile the notes from the conversations into organized clusters of information. It proved to be difficult to take anecdotal information and map it, as discussed above, but I tried my best to use more fluid organizers like word clouds based on the frequency that a particular issue was mentioned.

Style 3: Asynchronous Feedback
Stat: Around 8 small paragraphs received via direct messages, gchat and Discord

This was one of the least effective styles as it was clear that there was perhaps a lot of context missing or perhaps a lot of space for feedback to veer off course without live moderation. I was able to glean some edit possibilities from this style, as reflected in the website edit suggestions, but it wasn’t as rich as style 1 or 2.

Style 4: Website and/or Github Walkthroughs to Demonstrate Feedback
Stat: 2 conversations with 2 people

This style was a bit of a shot in the dark. I wanted to see if anyone wanted to walk through the website and the Github for me, pausing at all the places where they had meaningful feedback. I thought that perhaps it would help those less socially inclined relieve some of the pressure of conversation by using the website as a visual aid. I employ this tactic many times in the classes that I teach which is always beneficial for a subset of the population but I don’t think it was as eagerly adopted here. However, the two people I spoke to were user experience designers and their feedback was quite helpful. Their feedback is documented in my visual layout recommendation.

Style 5: Short Copy Revision Discussions
Stat: 1 session with 4 people

I think I might try to do one more session of this before the close of the Tech Writer position — it was very helpful and actually very fun! I think it would have been even more meaningful in person, the same way that an edit-a-thon can also be a convening of folks who share a similar passion for new or lesser documented sources of information. This session was focused on the community guidelines section with the lens of extending those values out to the rest of the website. Instances like this that allow for some play and positive social engagement around topics that are often very sterile help to normalize the incorporation of community guidelines into platforms, organizations and initiatives. Folks made some really exciting suggestions that are reflected in my edit of the website copy and layout.

A Quick Look at My Findings

Each website and github edit is articulated in a separate google doc, linked above. At the top of each doc I have provided context for my interest in investigating that page as well as a description of my findings in conversational language. Each doc then contains a copy paste of the copy that has been amended with coloured edits. You’ll see a detailed summary of the suggestions in those pages but here is a brief glimpse into some of the findings throughout my process. As a general review I was very happy (and not shocked) to find that several pages on the website are being received very well! Others could benefit from a boost in tone, a few tweaks, some related works and just a lot more context.

p5 Website: I have found through my interviews that people generally perceive the code of conduct and community statement to be static on one or two designated pages but aren’t confident that it is baked into the site at large. In multiple instances this has caused people to question whether the space is inclusive at large or just within a certain conversational context. I can understand this rationale — I myself look for many “green flags” as I call them that indicate to me that this is a space that holds itself accountable to the marginalized community that it references in it’s code of conduct. I think p5 is doing the work of building inclusive spaces in a very practical sense so it seems like there is room to reflect that more visibly in the organization of the copy.

In addition, it seems fitting to include some of the concepts in the code of conduct in the more technical pages. Why not “Get Started” by learning how to make your first sketch while also taking your first moment to consider the implications of your code? Why not explore these things in tandem? I’ve made some suggestions to the copy that prompt people to hold space for both the social and the technical — if we want people to be more mindful as coders we cannot approach the conversation in disconnected ways.

Github: These pages, it seemed, had a less positive review from the participants that I interviewed. I think that making welcoming content on github can be an uphill battle because the aesthetics of the interface are reminiscent of what is largely seen as white, male, exclusionary spaces. I think that this is tied to the reasons why platforms like p5 (as well as Arduino, Scratch, Teachable Machines and others) are designed to look nothing like this. They don’t have lines and lines of text that look like code, they incorporate design, they are colourful — all the things that might entice a new user to see this place as intentionally set apart from the old guard of technology spaces. The tricky part is aiming for those things while using an interface that has not diverged from that aesthetic. Participants and I had many interesting conversations about how the copy could aim to bridge the gap, including praising the ways in which it is already doing so in some places!

Additionally, the pages I investigated were causing new coders and new contributor participants quite a lot of stress as they had to open so many new pages to get the information they needed from one critical paragraph. I understand that too much copy all at once can also be a stressor too so I’ve made some edits that suggest, perhaps, a happy medium.

Very Brief Conclusion

I don’t think any of my suggestions are perfect but they are reflective of some of my initial assumptions, some debunked assumptions and most of the comments discussed in my ethnographic research. I don’t think work of this kind is ever done but perhaps this provides an alternative perspective on the existing approach.