Gabber: Capturing and making sense of audio capture for non-experts
Gabber is a digital platform designed by Open Lab to make capturing and analysis of audio recordings lightweight and accessible for everyday people and organisations who have limited resources available to do so.
There are two software components to Gabber: the mobile application allows people to collaboratively structure and capture an audio recording around topics they want to discuss, and the website is where participants of the recording can review and consent to how the recording is used, and comment on their conversations afterwards.
This study describes a process of iterative design with non-experts to create Gabber and identifies a need for flexible and informed consent processes to make these audio recordings accessible and meaningful to participants involved in the capture process.
Our paper illustrated that using simple techniques for capturing metadata familiar to non-experts (i.e. tagging) made this process inclusive, accessible and promoted in-depth discourse on the topics that were important to them.
Across three real-world deployments, Open Lab worked with university students and a local charity to trial Gabber, and iteratively refined the platform based on insights from each field trial.
How it works
Anyone can set up a project and can create topics to structure capturing conversations. They can set a number of discussion topics, which will be used as prompts in the app and a privacy setting of public or private.
Prior to recording, many participants can be added as being active in the conversation, which support accessing and modifying consent after creating the recording. When recording, a list of project topics is displayed to prompt discussion – tapping one of these indicates what people are talking about.
Once the conversation is uploaded, participants can listen to recordings and create comments directly on the audio to have discussions with other project members and (possibly) the public.
Curation and Content Reuse
Using conversations captured by the organisations and participants in our field trials was a key goal, however, selecting snippets of audio content requires time, resources and skills these non-experts often do not have or cannot develop given organizational constraints.
Our approach was to build on existing non-expert practices of listening, remixing and curating audio content: playlists. In our second final trial, we led a design activity to gain insights into how audio content was likely to be used, then the focus of our final field trial was to understand this playlist curation interface and its use in context.
Open Lab worked with university students and a local charity to trial Gabber, and iteratively refined the platform based on insights from each field trial. and it became clear that the platform could be used very differently across such diverse contexts.
Sensemaking of conversations
The sensemaking activities – allowing people to add comments to the final recording – allowed participants to reflect on their own experiences and understand how their existing practices affected others.
Some participants raised concerns about being reluctant to share opinion through comments, which prompted a redesign of the commenting interface to support pseudonymous avatars.
Across these field trials we found that individuals would take ownership over the configuration of the platform and directed participation from the other people involved. This became a bottleneck in the first two trials as the framing of discussions were driven by one person, which influenced the length and formality of the conversations captured.
This made us consider a more participatory approach and we wondered how to reconfigure the pre-capture phase to better engage people to give them more control over the types of conversations captured.
Further, the creation of discussion topics was fundamental to sensemaking and curation as the types of conversations (i.e. formality, length, etc.) dictated how participants could engage with the content.
Listening to one-sided or interview led conversations was less engaging for participants and the commenting practices meant that curating this content was more challenging.
Consent beyond data capture
Gabber was designed to democratize the collection, sensemaking and reuse of audio conversations, with the aim of making this content available more widely. This inevitably meant that rigorous consent procedures were critical for sharing this content.
Across our field trials, particularly working with the charity, it became clear that there was a need for flexible and dynamic consent processes. This emerged as our partner organisation wished to revoke consent on behalf of a participant due to their life circumstances, which indicated that our existing email-based approach to dynamic consent was insufficient.
Audio data is also not easily anonymised, which suggested a need for easily communicated guidelines about how widely the data should be disseminated (from an ethical perspective) and that should respect the participants consent choice.
Read the paper: Gabber: Supporting Voice in Participatory Qualitative Practices
JigsAudio is an open-source device that engages people in placeshaping and urban change by drawing and talking.
Our Story is a digitally mediated workflow for Participatory Video developed by Open Lab in collaboration with the International Federation of Red Cross and Red Crescent (IFRC).
We studied two Maker Faires to gain a better understanding of the complex dynamics that makers operate in.