CHI 2018: Strategies for Engaging Communities in Creating Physical Civic Technologies

Our second CHI paper this year is also from the Ardler Inventors project, picking up a little further down the line during the second and third stages of the project. The ultimate aim of the project was to explore how we could use hackathon-like events, which we called Inventor Days, to catalyse a community of people from Ardler in Dundee around using technology to support their local area. This built on our findings from previous fieldwork around hackathons.

We discussed the Inventor Days themselves in detail at DIS last year, but this paper takes a higher-level, end-to-end look at the different strategies used across the length of the project to support people in creating their own physical technologies for their local area.

You can find a pre-print online here and the final paper will be open access from the end of April.

Summary

We’ve been doing research with neighbourhood-scale technology for nearly 12 years now, working with communities like Wray, Callon & Fishwick and Byker. A key issue that emerged repeatedly through these projects was that, although situated and physical technologies like the Wray Photo Display or Viewpoint were positively received and played useful roles in the local area, these research prototypes often didn’t outlast the project they were associated with.

Instead, our latest work has tried to understand how we can create skills, enthusiasm and relationships that would help a community to identify new uses for technology in their local area, prototype ideas and try them out themselves. This paper describes our experiences of designing and testing a number of strategies:

Inventor Days. These hackathon-like events were the backbone of the project, which aimed to bring together residents from Ardler with members of Dundee’s creative and maker communities. Through a series of three events, residents and makers explored the neighbourhood, developed ideas for their local area and prototyped early versions together. One of the most successful approaches at this stage was re-orienting the events around local knowledge rather than technology, casting attendees from Ardler as the experts.

Inventor Kits. For the next stage, we took ideas generated through the Inventor Days and created kits that could be assembled and modified by community members. This was intended to explore how ideas from one neighbourhood could be replicated and built upon other communities, and how the provision of simple physical prototyping components could support creativity. Something that worked well at this stage was the perception of the kits as a kind of gift, which respected their earlier efforts while allowing them to create a much more finished version.

Community-Led Deployments. Finally, we wanted to understand what challenges participants would face in actually deploying and testing the things they had designed. It was important to both us and the participants that the things that they built actually saw the light of day, even in some temporary form. Alongside the practical difficulties, one of the main things we noticed at this stage was the importance to participants of showing their work to others and gaining feedback in controlled settings, whereas we pushed towards full deployments to test the ideas.

Key Findings

The paper has a lot of practical lessons from each of these stages, some of which I’ve touched on above, but I’ll focus here on some of the higher-level emergent themes in the results that will be most useful to people trying to engage communities in this kind of activity.

Motivations. Going into the project, our framing was largely about solving problems, but community members never responded particularly strongly to this. While most neighbourhoods have their issues, few immediately lend themselves to the type of technologies we were using. Instead, their motivations were about developing skills, expressing themselves and doing things together as a family. As we found during the deployments, the rewards they gained were similarly centred around personal development. This suggests to us that HCI needs to move beyond exclusively framing neighbourhood-level applications of technology in terms of problem-solving.

Ownership and investment. At every stage of the process, we tried to install in a sense of ownership over the prototypes, and it seemed like participants felt real pride in them. But in hindsight, it’s clear that they didn’t have much ownership of the process itself, which was very much driven by us. It’s clear that some degree of bootstrapping is necessary, whether that’s from outside or within the community, but we needed to be more active in ceding increasing control to the community at an earlier stage.

Scaling and maintaining enthusiasm. Finally, our goal was to create something sustainable, and while there are some signs of ongoing activity, this proved to be very difficult. The tight-knit group of core participants was a hugely positive experience for us, but it also made it difficult for new people from the neighbourhood to join in later. As well as thinking about how this core group of participants can be supporting in taking ownership and leading the process, we also need to think about ways to make that process more transparent to others in the neighbourhood and support more lightweight types of participation.

British HCI and HCI in Britain

Note: I started writing this last month to capture a lot of different conversations I had at British HCI (and which I think everyone else was having too). Long before I bothered to finish it, John Vines distributed offical notes from the town hall meeting, which are much more complete and well worth reading. But they don’t have an anecdote about Greg Abowd.


British HCI was the first conference I attended as a Masters student in 2007 (I didn’t talk to anyone). I attended again in 2011 when it was held in Newcastle, but last month in Lincoln was the first time I’d been when it wasn’t within walking distance of my flat. I think it’s fair to say that the conference is not held in high regard—or at the very least that it has failed in recent years to attract a significant chunk of the HCI community. If I remember rightly, Greg Abowd’s closing keynote in 2011 was itself less than complementary.

This year’s conference, hosted by LiSC, openly sought to reenergise the conference. I don’t think anyone who attended would deny that they achieved it.

What struck me most was that the papers were the least interesting thing on the program. The ~28 full papers were dwarfed by by position papers and invited talks, panels on the REF, designing with data and civic engagement, a research council lunch for early career researchers, an unconference lunch and no less than four keynotes (all from outside HCI). In particular, everything I heard about the doctoral consortium was glowing—some 50 students in a workshop that actively engaged them, rather than half a dozen presenting their work to each other.

Back in 2011, I’m pretty sure Greg Abowd’s point was “stop trying to be a shadow of CHI”. While some regional conferences have managed to make this work and carve out a niche (NordiCHI, for example), this was clearly not working for the UK. At the time, it wasn’t clear what else the conference should try to be. After this year’s conference, I think it’s becoming evident.

What should British HCI be?

What HCI in the UK really needs is better networking on home soil. It’s insane that there are many people in the community that I may only ever see abroad. For whatever reason, British HCI wasn’t bringing people together, and I’ve often thought that the annual Digital Economy conferences were better networking opportunities (particularly given that the largest HCI groups were holding big Digital Economy money and were somewhat obliged to attend or host). Notably, Digital Economy papers aren’t archival.

Making the rest of the program so interesting allowed the organisers to drop the pretence that the papers are important. Instead, they were able to treat the conference as an event to bring people together, share what we’re doing and discuss hot topics. As a result, it had a very different (and better) feel.

But you can’t fix everything at once. Cost remains an issue—at £600 plus meals, it still wasn’t exactly cheap for an early career researcher with no budget. There is only so much you can do about this, but it’s something that the organisers will need to take into account. Heavily subsidising PhD attendees and providing very cheap accommodation was a positive step and allowed us to send some of our students. This sort of thing should be a minimum requirement for anyone who wants to host it.

It’s also tangled up in the issue of who should champion HCI in the UK. Like the conference itself, BCS’s Interaction Specialist Group (which normally endorses the conference, but this year did not) has been marked by a lack of engagement with much of the HCI community, particularly the groups that have emerged as powerhouses in recent years. In the town hall meeting, I pointed out that SICSA is already doing an admirable job in Scotland of many of the things we might want a UK HCI body to do—mainly networking, but also showing leadership on key themes, engaging with industry and funding activities. That said, Scotland has a number of properties that makes this much easier: most of the research groups are clustered within an hour or two’s travel and SICSA is backed up by funding from SFC. Scaling this kind of activity UK-wide might be tricky, but it’s certainly worth looking at what they’re doing.

The meeting didn’t reach a firm conclusion and it’s an ongoing discussion that we need to keep having. Likewise, the conference needs to keep innovating. 2015 doesn’t necessarily need to be the template for future conferences (in fact I’d argue that the distinct local flavour was to its benefit), but it should be a positive example. I don’t know if I’ll attend next year (you couldn’t get much further from Dundee than Bournemouth), but I hope some of the momentum is carried over.

Mea Culpa

PS. I don’t fully remember how I ended up with this, but I hope it didn’t cause any problems for the closing plenary. Whoops!

IMG_2943

TICTeC 2015: Revisiting the Myth of Digital Democracy

The Myth of Digital DemocracyI recently read Matthew Hindman’s The Myth of Digital Democracy [1], in which he tears down the notion that the Internet has significantly democratised participation in the political sphere. I particularly enjoyed his data driven, economics approach to the issue, which reminded me of the methods used in Freakonomics [2] and my very favourite example of counter-intuitive cause and effect, in which postal voting actually decreased turnout at Swiss referendums [3].

Hindman’s core argument was that structure of the web itself prevents it from truly democratising participation—the way Google works, for example, bestows a rich-get-richer effect on content that is already widely linked. Interestingly, though, his work was based on a pre-social media Internet and he was largely talking about blogs as a mechanism for political discourse. I was left wondering what effect social media—which has undoubtedly changed some of the structure of the web—might have on his findings. Sharing and retweeting, for example, are recent structural elements that provide new opportunities for discovery (but certainly don’t even things out!)

With a focus on the impact of civic technology, and on web-based civic technologies in particular, it was natural that these issues would crop up at TICTeC. Shelley Boulianne’s opening keynote addressed it head-on with a meta-analysis [4] of research examining whether Internet use has a positive or negative effect on civic engagement. You don’t come across a lot of meta-analyses in design-led HCI research, so this was a nice change! Her analysis of 38 studies showed an overall positive effect. What’s interesting in this case is that it’s not the Internet acting as a tool for participation, but rather to better inform citizens and inspire them to participate through more traditional means.

Ethan Zuckerman’s closing keynote likewise did a great job of dismantling some assumptions around slacktivism—low-effort online activities like putting a Twibbon on your profile picture, which are widely regarded as being ineffectual. He argued that these simple but visible actions can actually go a long way towards challenging and ultimately changing social norms. For example, if a large number of your friends and family are visibly supporting gay marriage, this has a lot of normative influence. In this example, it is not just social media’s ability to get a large number of people involved that makes it effective, but also its ability to draw on influential social ties.

I think these examples go a long way to illustrating the wide variety of ways in which technology can have an impact, but also that sometimes need to go digging a little deeper to find them.

  1. Hindman, M. (2008). The Myth of Digital Democracy. Princeton University Press.
  2. Levitt, S.D. and Dubner, S.J. (2005). Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. William Morrow.
  3. Funk, P. (2010). Social incentives and voter turnout: evidence form the Swiss mail ballot systemJournal of the European Economic Association 8(5), 1077–1103.
  4. Boulianne, S. (2009). Does Internet use affect engagement? A meta-analysis of researchPolitcal Communication 26(2), 93–211.

TICTeC 2015: Take-Home Thoughts on Civic Tech

TICTeC-logos_general-with-yearAt the end of March I attended the first TICTeC conference, run by mySociety in London. If you haven’t come across mySociety yet, you will almost certainly have come across one of their websites, including TheyWorkForYou, FixMyStreet and YourNextMP. They now have a research programme aimed at understanding how effective these tools are, part of which includes running an annual conference bringing together people interested in civic tech.

I turned up to give a presentation about Viewpoint and I don’t think I’ve ever been to an event that seemed quite so tailored to my interests—in fact, I think I’ve run events I was less interested in. There was a lot to soak up, but now that I’ve had a couple of weeks to digest, here are some take-home thoughts.

How do we measure the impact of civic technologies?

Opening the conference, mySociety CEO Tom Steinberg set out the importance of this question with a neat analogy: in the past, doctors prescribed all sorts of things that we now know to be hokum, because they lacked the ability (or will) to rigorously test their remedies. If we develop civic technologies but don’t rigorously test their impacts, we are effectively doing the same thing.

When talking about Viewpoint I’m always candid about the fact that while we succeeded in our goal of generating high levels of participation, with hard data to prove it, we failed to find a way of converting that participation in to meaningful outcomes. What’s more, we didn’t really have a clear idea of how to evaluate that. It was both reassuring and concerning to discover that most other people working with civic tech didn’t really have a clear idea of that yet either.

By coincidence, this question has emerged on three separate projects I’m involved in over the subsequent couple of weeks. One of the solutions mooted has been measures of self-efficacy or collective efficacy that measure people’s belief in their ability to effect a change before and after the intervention. This sounds quite neat and tidy, but I’ve never quite been convinced by the simple ten-point checklists that have been developed for this purpose. Slightly more convincing are community-level indicators, which measure conditions in the community tailored to your goals. However, both these approaches are confounded by the time-scales we work with: research projects are increasingly tending towards the 18-month variety, but these effects can take a long time to emerge.

What is the value proposition to civic organisations of getting involved with researchers?

In contrast to most conferences I attend, practitioners vastly outnumbered academics and it was interesting to see how the non-academic civic tech community perceives us. The two communities have a lot to offer each other, but articulating the academic contribution and working around practical challenges is difficult. It’s something I spent a lot of time banging by head against while working on the Creative Exchange (with only a headache to show for it). Not least amongst the practical difficulties is the seemingly glacial pace that academia works at. Even our shortest projects don’t mesh well with small organisations who need to be fleet of foot.

There’s also the issue of money: we have a bad reputation for assuming organisations will be willing to take part in research projects for free. It’s not entirely our fault, as RCUK is often the most attractive source of funding for us, but typically won’t hand anything out to non-academic partners. There is some sign of change on this front: In the Making is funded through a strand of AHRC funding that allowed groups representing communities (in our case Disability Rights UK) to be funded as Community Co-Investigators, and the AHRC Knowledge Exchange Hubs are also handing out limited amounts of funding to non-academic organisations. I’d love to see EPSRC’s Research in the Wild strand move in this direction, where it seems particularly appropriate.

How do we think beyond the web?

My last thought is more of a self-centered observation. The vast majority of work I saw presented was web-based and much of it was based on national-level discourse. There were exceptions: Nanjira Sambuli spoke about the importance of local radio and it was great to meet some people from Code for America and their international siblings whose work can potentially take all kinds of forms. But for the most part of idea of civic technology is currently grounded in the web.

The web is obviously pretty great and a powerful tool for civic engagement—you only have to look at the level of discourse going on right now around the General Election to see that. But if we fixate on the web, we risk being blinkered in the same way that Old Media was. Civic life works at a lot of different scales and the web isn’t suitable for all of them—at least, not the web as we currently conceive it. I went along hoping to show that we can think much more creatively about the role that technology can play: I don’t know how effective I was in doing that, but I think it’s something that bears repeating.