FOSS Interview Workflow

17 Apr 2022 - Robert W. Gehl

With my last book wrapped up, I am now hard at work building the next book, which is the Goal 2 of this blog. (Goal 1, recall, is writing about being a FOSS academic. Goal 2 is writing a book about FOSS). My new book project will focus on what I’m calling “ethical source alternative social media” – social media that allows ordinary people to be in control, that uses moderation for social, economic, and environmental justice and not to support oppression, and that runs on FOSS.

Methodologically, this project will demand I use my expertise in software studies, science and technology studies, and critical communication studies to analyze code, protocols, and network structures. But I also want to use participant observation and interviewing. And, I want this to adhere to the highest ethical standands as well as use FOSS technologies throughout.

I want to include the voices of ethical ASM members, developers, and admins. It may be tempting to do what a lot of people seem to have done and quote fediverse posts. However, I’ve observed that many, many fediverse users do not consent to having their posts taken out of context. There are public forums I can quote – for example, the Github issues forum for Mastodon. But I think the best vehicle to hear people talk about their experience with ethical ASM is via interviewing.

In this post I will briefly describe how I will manage interviewing. I hope this resource is useful not just to other researchers, but also to anyone considering being interviewed by me who might be curious about my process.

Step 1: Participant Observation

My first step has been going on for a long time: I’m participating in ethical alternative social media (and quite a few systems I would not categorize as ethical). As I do so, I take notes on the discourses happening on these systems, and I try to figure out which participants are passionate and invested ones. I want to talk to developers, administrators, and members of ethical alternatives. I note who they are with the intention of inviting them to participate in interviews.

Step 2: Inviting People to a Conversation

After coming up with a list of interesting people to speak to, I reach out to them. Most likely, this will be via a fediverse DM or email. This is the moment when I explain my purpose in talking to them and how their ideas can contribute not only to the book project, but also to magazine articles or academic journal articles. I can also ask them to look at an informed consent form – a vital part of ethical research practice.

I invite people to talk via whatever their preferred medium. If they have no preference or are comfortable with Jitsi video chat, I have the Jitsi plugin for Nextcloud ready to go.

Step 3: Preparation

To be ready for the interview, I should read/watch/interact with as much material from the interviewee as possible. I look at their public posts, blogs, artwork, videos, writings – just to get a sense of who they are and what they care about. Since so much of FOSS culture is pretty technical, I pay a great deal of attention to technical details so I am a bit more informed before we talk.

I write out about a dozen questions, ranging from asking for names/pseudonyms and pronouns to open-ended questions meant to spark a conversation.

Step 4: The Interview

If we are using my Jitsi app, I can record the interview (assuming of course the interviewee consents not only to talking, but being recorded). While it will record video, I am more concerned about getting audio. Jitsi’s Nextcloud app has one flaw: it records to Dropbox. I am not a fan of this but I can’t work around it just yet. (If anyone has advice about how Jitsi recordings could be routed to Nextcloud, I’m very interested).

I go through my questions, but in my experience conversations don’t always lend themselves to sticking to the script. That’s just fine – I use my questions as a general guide, but the discussion can go in other directions as the interviewee leads.

While I interview, I take notes on paper – after all, what if the recording fails? Paper notes also come in handy for the next step.

I will conclude the interview by asking if the interviewee has questions about me or my project, other things to add, and if they can recommend other people for me to talk to.

As soon as the interview concludes, I remove the recorded file from Dropbox and store it on my Nextcloud.

Step 5: Journaling

After the interview is over, I write a journal entry with my thoughts on the conversation – things like good stories, new ways of thinking about the topic, and a general summary of the conversation.

I am also happy to share my notes or the recordings with interviewees, so I follow-up via email to mention that and thank them.

Step 6: Transcribing

This is an area where I am tempted to use automated systems, but must refuse. After all, most transcription services, like Otter.ai, require me to upload audio to the cloud. No thank you. I think that violates confidentiality, and it also violates my desire to have this be as FOSS and self-hosted as possible.

So, this is the stage where I either transcribe or at least take very detailed notes on the conversation. Full disclosure: I may start collaborating with a grad student; if so, they may get involved here at the transcription stage. But for now, it’s just me doing this work.

Conclusion: Critical Reflection

So, that’s that: the interview process. It’s almost entirely FOSS/self-hosted, with the exception of Jitsi recording to Dropbox. And I hope all of this is transparent and ethical.

One practice I engage in throughout the above is: can this be better? For example, does my workflow adequately protect interviewee data? Am I doing a good job recruiting interviewees? Are my questions prompting a good conversation? And can I get away from Dropbox?

And, what am I missing here? If you have comments on this process, please let me know via the Fediverse comment system below.

Post Tags

• Goal 2
• Goal 1