This is the first in a series of posts detailing the lessons that the Ada Lovelace Institute, Traverse, Involve, and Bang the Table team are learning as part of the online deliberation we’re running together on COVID-19. For more information on the project – the introductory post is here.
We'll be updating this page weekly with lessons from each stage of the project.
Designing a rapid process: lessons from the setup phase
Our main challenge has been refining and confirming the research questions we will explore with participants.
- How do you design deliberation on an ever-shifting subject matter? We’ve talked to a number of people about the best focus for this work to make it as useful as possible and to build on other work being done in the space. However, things change daily, and, for example, we’re not sure by the time we are talking to participants what the status of the NHS app will be, whether the lockdown will have been lifted to some extent. In response to this, we have decided to start broad, and use this as an opportunity to give participants some autonomy over the process by asking them to prioritise the subjects they would like to discuss.
- Is a subject still worth discussing if the news cycle has moved on? There feels a great responsibility to be as current as possible, but just because we’re not talking about (for example) symptom tracking as much as we were when the ZOE app was launched, that doesn’t mean the issues have gone away or that it’s not relevant. We’re trying to ensure that we focus on the long-term implications for citizens, and some of the big questions that surround many of the proposed intervention to ensure the work remains relevant.
- How do you do something quickly and thoroughly with a constrained budget? Well, with some difficulty. Firstly, we know this work will have its limitations and we will be as explicit about that as possible (expect a blog on that coming shortly). Secondly, we’re assuming that even if it doesn’t seem that much other similar work is occurring, we know that there is – take, for instance, the Sciencewise Sounding Board pilots in 2015, the efforts to move the Climate Citizens’ Assembly online, and that insights from that work and other more emergent work shaped by lockdown will also help test and situate our findings. Thirdly, we’re trying to collaborate as much as possible to use feedback and challenge from partners and friends to make our work more robust.
Running online discussions: lessons from week one
We are running this project as rapidly as possible, and one consequence of this is that we are limited in how strict our sampling criteria can be. Compared with the sortition approaches of projects like the Citizens Assembly, our approach of an open invite to local online forums and groups is fairly crude. We are by default only including those who are online, although by using local groups set up specially to support communities during lockdown we’re hopeful this is a broader pool than it might have been three months ago.
That said, within about a week we managed to recruit a sample that within spitting distance of national levels for age, ethnicity and approximated social grade, split across rural and urban populations. Interestingly the one criteria we struggled most with was gender, even after snowballing from our initial respondents (e.g. we asked them to help us recruit more male participants). This isn’t an unusual finding, but it is likely that the method of recruitment (mutual aid groups, local resident forums) leans heavily female and this needs to be managed actively from the start of recruitment.
Logistics of zoom rooms:
In our first online session we had 26 participant, plus facilitators, notetakers, speakers and observers, bringing us up to around 35 people in the intro session. At this scale the chat function can feel a bit frantic, but it worked well for information giving. As soon as we broke out into smaller groups (around 6-7 people) the dynamic was much better and easier to facilitate. This was fairly easy to manage in Zoom, so highly recommend it.
Also – that thing where participants show up early? That’s true online too, your waiting room will be busy so be aware of that!
We’ll say more about this in subsequent weeks as the groups spend more time together, but initial feedback from our facilitators was that some of the usual dynamics (e.g. more dominant personalities getting more air time) were strongly present and a little trickier to manage online. Some of the tools we use as facilitators to manage discussions, from speaking more quietly to model a calmer discussion, to how we physically arrange a room, aren’t immediately available online.
We’re also finding that the process of establishing a productive discourse takes more time. we had originally planned one hour discussion sessions, but this is just too short, for us and for our participants – we were just getting going. And too much structure can get in the way of the discussion, and takes longer to explain online, so keep it simple.
We welcome feedback, guidance, critical challenge, and support – so do get in touch if you are interested in having a conversation about this work. Drop Anna a note at email@example.com