Communicating statistics is challenging, particularly for a national statistical institute. The demands of statistical accuracy, the challenge of communicating uncertainty are seemingly diametrically opposed to the need to talk simply and clearly to mainstream audiences.That was the challenge in front of the Office for National Statistics when it began work to consider how it could meet the needs of citizen users and launched the Visual.ONS website. In March this year the Visual.ONS Beta draws to a close as three years’ worth of innovation is migrated to the main ONS website where the content aimed at a broad audience will find a home. It is the culmination of a three-year experiment that has changed the way the ONS communicates. But first let’s set the scene – in 2015 Visual.ONS was launched to help develop the ONS’ communication skills by wholeheartedly embracing digital story-telling – combining the talents of statisticians, analysts with data journalists, designers and data visualisation expertise to help a broad audience connect with our data, to understand change told through innovative formats they recognise and embrace. User research had uncovered a large audience being under-served through our existing statistical products, and the ONS had been encouraged by both Better Statistics, Better Decisions and the House of Commons Public Administration Select Committee to make positive strides in how communicated.The latter’s report said: “Government statisticians should work much more closely with different kinds of users in order to present statistics in ways which meet their different needs.”Improving the statistical communication for a broader audience required a fundamental change to the way we commissioned, produced and published our content.We had to develop and hire the skills needed to improve communication – from data journalists to data visualisation staff, and training designers and social media staff. In three years the Digital Content team has been almost entirely hired from outside the ONS, with significant investment in training and development of those ONS staff who were on the project from the start. Visual.ONS has been built on a collaborative relationship with statisticians and analysts – where ideas are shared, projects are co-developed and skills and knowledge is exchanged.Visual was designed to meet the exacting standards of modern digital products – responsive across all device types, accessible to meet the needs of all users and shareable so that 3rd parties can embed and make use of the content created.In three years we have covered topics such as helping users understand life expectancy, causes of death, government welfare spending, the impact of Brexit, decades of social change and the gender pay gap.
What makes this project innovative?
The innovation is in the collaboration - taking the work of statisticians more used to expressing their work through methodology papers and putting them on projects side by side with editorial teams to find new ways to express the data. The biggest change and innovation as a result of the work is finding a shared vocabulary within the organisation where statistical precision can still be married with messages that work for a broad audience.By focusing on topics driven by user questions, helping users explore data through personalisation and interactivity and through language and form they are used to seeing on media websites we’ve been able to see ONS data travel further, deeper with audiences.The work of Visual has strengthened the editorial focus of our communications, with a greater focus on commissioning on the basis of user need, through a strengthened governance process.The team has produced a range of innovative story-formats that have resonated with users. From a social exploration of the changes to marriage data, to helping users explore changes in the economy since the EU referendum Brexit vote and unlocking dense spreadsheets through smart explorers these formats have been powerful new ways to communicate statistical information.
What was the impact of your project? How did you measure it?
As the Beta ends, Visual can be said to have had a transformative impact on how the ONS communicates to audiences.The Digital Content team has worked with statistical colleagues to produce 200 separate pieces of content in 3 years, as well as thousands of accompanying bespoke social media graphics. Almost 2.5 million people have read Visual articles on the site since launch, while more than 6 million people have consumed the maps, tools, interactives and graphics created for Visual because they’ve been embedded on the websites of publishers such as Guardian, the BBC, the Daily Telegraph, the Daily Mail, Business Insider and many others.We’ve reached a different type of user too – a far higher proportion of people reach Visual content via social media than outputs on the ONS website, for example. Engagement on Visual stories is 40 to 50% higher than on statistical bulletins in terms of dwell time, while the viewership is typically 5 to 7 times greater than articles on the ONS site.We’ve collaborated directly with internationally-recognised media partners, such as The Guardian, BBC News (we’ve included one of the collaborations as a chosen link) and Buzzfeed: the ONS editorial and statistical teams working directly with journalists at those organisations to produce content which helps cut through to audiences with clear statistical messages.The BBC collaborations - Risk of Crime and BBC Today: 60 years of change - reached more than 400,000 and 2,000,000 people in a matter of days - greatly extending the reach of ONS analysis.The work of the Digital Content has been celebrated by our peers, and by our users – through scores of comments on social media, to wider national and international impact.In the last three years our digital, statistical and editorial approach has seen us share best practice with Eurostat, Stats New Zealand, NHS Scotland, among others.Late last year our work was also cited in a recent Welsh Assembly report on Digital Dialogue, which recommended putting the user at the heart of its communications.It said: “The Assembly must exploit all alternatives to the press release as a means of promoting its work. Maps, infographics, blogs and neat summaries all have the potential to articulate difficult messaging in a memorable way. The Office of National Statistics’ visual.ons website is a particularly good example of how this can be achieved. Having understood the breadth and depth of information and briefing materials generated by the Assembly’s research service, we recommend that this data-rich service’s outputs are included as a core part of Assemblynews and information content packages in future.”
Source and methodology
The source of the overwhelming majority of data used is, of course, the Office for National Statistics.The methodology varies from project to project but there have been some innovative approaches:In the case of the risk of crime calculator we combined location, personal characteristics data (age, sex, housing status, occupation status) with deprivation data and three years' worth of crime survey data to produce a model which determined individual risk by crime type.More information on the methodology is here: https://www.ons.gov.uk/visualisations/riskofcrimemetadata/index.html
The team largely uses d3.js to produce visualisations. But in the case of tools such as the Draw Your Own Charts it was D3.js (with a few supporting libraries to make it work – pym modernizr etc). and then effectively hand coded. We used a starter that was made public by the NYT – but with significant changes (draw in reverse, styling, handle different date formats, driven by a config file etc)
Rob FryFrank DonnarummaZoe HartlandLaura HardingHenry LauJure StabucJohn NixonLisa JonesPhil LeakeSophie WarnesCallum ThomsonAnd with the BBC collaboration:Wesley StephensonJohn Walton