On Saturday I caught an early train to London to attend the latest UKGovcamp. If you don’t know what that is then the event describes itself as a;
“..self organised unconference for people that work in and around government.”
Though that does rather assume you know what an unconference is [check out the Wikipedia definition.]
Basically Govcamp is made up of a great deal of people with an interest in digital and the public sector all giving up their Saturday, lots of available rooms, decent wifi, a fair amount of coffee and *a lot* of talking.
This year it was taking place at the heart of London’s political establishment – City Hall. It really was an amazing venue to host this sort of still slightly subversive event.
This was my fourth and it is always a highlight of my year. This year was a bit different though as for the first time I found myself co-leading a session – something I have always scrupulously avoided in the past!
Jeni Tennision from the Open Data Institute pitched a session around the topic of ‘statistics on the web’ and much to my surprise it proved pretty popular (there was a great deal of competition for peoples attention). In fact so many people (40ish) showed up for our session that we needed to swap rooms as we didn’t have space.
As you would expect with such a crowd the talk was pretty technical and very focused on the open data aspects of what the ONS does. It was pretty eye opening though.
It was clear that many people see the ONS as a creator of standards in this space almost by default and feel we have a responsibility to do a better job sharing these underpinning elements – particularly classifications/identifiers that would make it easier to compare datasets across providers swell as our own.
There was clearly a demand for more properly machine readable data – whether via an API or simply well structured CSV at persistent URLs was up for discussion. The lack of awareness about the work that ONS has been doing around the ‘web data access [WDA]‘ programme, including the data explorer and API was interesting given many of the people speaking were particularly interested in the very Census data it is currently providing.
Many of the questions and complaints people had could at least be partially answered by the functionality that WDA is starting to offer so that was heartening.
It was however clear we have miles to go when it comes to usability issues. It was clear that even beyond the standard UX and findability problems we are well aware of that even some quite expert users find the usability of our data and what they can and cannot do with it a struggle. More than once on Saturday I heard the phrase ‘context is king’ and there was a real demand from this room at least for us to offer much more context around our datasets. The history of data sets, what they link to, the methodology behind them, suggestions of what can be done with them and what they can safely be combined with. This group was even interested in what tools could be used to manipulate the data beyond just Excel and perhaps some recipes on how to achieve that. Most if not all of this information is available but not easy to find or easy to comprehend without a deep understanding of statistics.
The ‘linked data’ work Ian Coady has done for ONS around Open Geography was universally praised which was another nice spot and Ian deserves all the kudos he is getting.
I want to thank Jeni for proposing the session (even if I wasn’t very gracious about that at the time!) and for everyone who came along to it. It was really useful and it certainly gave me a great deal to think about.
Elsewhere at the event I was getting a very different perspective of statistics on the web. The new web experiment from the Mirror newspaper, Ampp3d, was a frequent topic of conversation. This idea of a (mobile first) website that focuses on the headline numbers from big news stories and then digs a little deeper has clearly struck a nerve. There is no doubt Ampp3d serves a particular political agenda but I was able to introduce people to some of the work Full Fact do for something similar with a more ‘neutral point of view’. I think the technical focus during the session and these corridor conversations show just how wide our continuum of potential users is. Again it gives me a lot to consider.
In fact the entire idea of ‘data storytelling’ was a significant theme this year to the extent that a couple of us are considering a more focused spin-off event on that very topic. More news on that coming soon.
It was a brilliant, thought provoking event and I have nothing but praise for the volunteer organisers. Running an event at this scale is not simple when it is your day job let alone when you are doing it in your (limited) spare time.
So thanks and see you next year!