A few years ago, our session host, Rory Gianni, through being involved with several open data initiatives, saw that some went on to great success and some weren’t sustainable. One factor that seemed to make a difference was engagement – if you are not involving people outside te organisation, why are you doing it? Even if you’re being driven by the stick of legislation, you could still capture why others would be interested.
He has a set of digital engagement notes on GitHub. These follow on from the five stars of open data engagement, conceived at UkGovCamp in 2012.
You need to make any data you publish both findable and accessible. Would doing both a spreadsheet and a fancier format make it more accessible, for example?
A bare minimum of engagement is that somebody should be able to give you a correction if something is wrong. Too many people think of publishing as a one-way process, rather than one with feedback. A contact form — or even a forum — could give users ways of communicating about problems with the data. Sometimes the data provider doesn’t need the one responding – others can support each other. If the publishers don’t do it – the user can get together and do it theirself.
Stories of people doing things with the data can be compelling in developing engagement. For example, people found unknown Roman Roads via LIDAR data. But it’s hard to find out how people use the data, unless they come to you. Can we make that easier?
Consultation isn’t always engagement — it’s a tick in the box, rather than a genuine dialog.
Why didn’t sites like GitHub and SourceForge work in the data space? There are a mix of reasons – software development has more of a history of openness, and there’s a bigger split between the publishers and the users in data. People are usually both users and publishers on GitHub. Software forks more naturally.
Data needs a mediation layer – like data journalism – that make sit accessible to non-specialist consumers. Should we be thinking about more mediation layers? Data journalists often struggle to find the data sources they need, but they do talk amongst themselves in various ways.
Motivation for data use
You don’t need to be an expert in media, or data to make use of data – you need a motivation, and an interest. Do publishers think about the communities that are most likely to use it? Who would be the first person to notice a mistake in your data set? There’s some resistance to these ideas — “open data is for everyone” — but it’s hard to engage with everyone, and you can engage with the most likely communities to use it more easily. Can you behave more like your consumers – and test your data and the way you publish against that?
Can you give people the first couple of steps to get going? Perhaps point them towards a tool that can read the formats you publish? Then the support transfers to the tool providers or user community?
From a school level, children start being trained not to ask questions, to accept answers. We need to break that, and have a more collaborative approach to problem solving and learning. Finland has quite high levels of resilience against misinformation, because they train their children to assess the source of data. In the UK, we’re behind in the content of the curriculum. There’s a societal health issue here – the data illiterate are ill-equipped to navigate the world.
Shall we let data be taught as a skill, not a subject? #odcamp
— Katya B (@PitbullPossum) November 4, 2018
Contextualising the use of data as a toolset rather than a “subject” might be an answer – see the story above about LIDAR data proving to be useful for history and archeology. Gathering these good news stories really matters, particularly ones that express the benefit that was obtained.