Most engaging presentation of local area research
Oxfordshire County Council: Creative responses to local Data Users’ needs: Data Stories and R-Area Profiles.
Oxfordshire County Council has developed new Data Stories following a consultation with local data users. Data Stories provide data on a topic as a sequence of interactive infographics. This enables novice data users to access the data on a topic in a structured, simple-to-use and understand way. Unlike conventional static infographics, Data Stories let the user customise the data contained in the infographic for their individual needs. Presenting data this way makes is more accessible for the average user, who struggles with statistics, and also makes it possible to access data on mobile devices which may not handle spreadsheets well.
Some users requested a written report at a more local level rather than a series of webpages – particularly data users working on Community-Led Plans and Neighbourhood Plans for their Parish. To meet this need for very local data in a report format, we have developed automated PDF report production using R: some R code accesses open data sources by API, prepares graphs and tables and updates text. This will, for the first time, make it feasible to provide a PDF-format profile for every parish in Oxfordshire, giving local data users the data they need in the format they requested.
To be effective, Local data provision should be structured around the needs of the audience first, rather than focusing on advance technical solutions. What is important to analysts is not always what our data using “customers” need.
Oxfordshire County Council’s Research and Intelligence Unit’s longstanding data website is aimed at local partners and the public. Recent years have seen huge changes in data consumption, with people now expecting to access data from the web quickly and easily on a smartphone, expecting rich information without needing to download underlying spreadsheets.
Reviewing the requests for data received by its Enquiries service, the Research and Intelligence Unit were increasingly aware that users were requesting basic information (such as the population of the County) by phone or email, rather than self-serving from the website. This strongly suggested that the website was not meeting the needs of data users. It was vital to address this, not least because dealing with regular in-person enquiries for basic data is unsustainable.
Research and Intelligence surveyed local data users in summer 2015, in partnership with other groups, to understand users’ needs. The survey asked about data they most needed and their success in finding it. This gave us valuable evidence about the needs of local data users.
The team, jointly with Communications staff, underwent training in Infographics development, seizing the opportunity to present the data that local data users need in a creative way, taking advantage of our growing confidence and proficiency in the Tableau software we had procured several years before.
At the same time, some of our customers for very local data (those preparing Community-Led Plans) wanted static area profiles. Producing profiles for each of the hundreds of parishes in Oxfordshire manually would not be feasible within the resources available, so our more programming-oriented team members investigated using APIs using R. We quickly realised that we could develop automated Parish Profiles in R, complete with customised text, graphs, and tables.
The first result was Oxfordshire’s Data Stories: topic-based collections guiding users through the key points and simplifying access to customised infographics and, when required, the underlying data. The structure of the stories means that users first get an overall infographic summary of the main data and the key messages, rather than being dropped into a “sea” of confusing, detailed data. The “chapters” of the stories enable users to dip into more detailed data sub-topics when they are ready. Like standard infographics, Stories present information in a simple, uncluttered way: but are also interactive, responding to icons that users can click or tap to customise the infographic to their needs. They are designed to fit perfectly onto mobile device screens without compromise, and can be easily updated with new data, or with additional chapters, making this a sustainable approach.
The second result was automated Parish Area Profiles: using a template built in the R language, a profile containing text, tables, and graphs is prepared automatically in seconds. This speed makes it feasible to produce Profiles for multiple parishes. The data is sourced via application programming interfaces to national datasets, meaning that the data contained in each Profile is always the most recent available at that moment.
Data is no longer consumed only by data experts: the public is used to consuming data through apps and webpages. However, simply providing data as a spreadsheet or table doesn’t meet the needs of most users. We found that:
- When there is a lot of data on a website, such as Oxfordshire Insight, people can be blind-sided by too much choice and information. A vast “sea” of data is unhelpful (how will they know this figure is the key figure for this topic or the latest table?);
- Many users of “very local” data want data in a report format, not online pages or dashboards.
- As few as 15% of people have studied maths post-16, so we need to design data tools for the needs of the non-data-proficient.
- People using mobiles and tablets may struggle to download and use spreadsheets of data.
Local data provision should be structured around the needs of the audience first, rather than focusing on advanced technical solutions. We think we have demonstrated creative new ways to meet user needs while making use of open data and free (Tableau Public) and open (R) software, and whilst preparing for a low-resource future.
Find out more
The first of our Data Stories are available from the Insight homepage: these cover Population and Birth, Death, & Migration, which were the most asked-for topics. Also available from Insight is a link to the survey of local data users.