ODI data literacy report laments technical skills inflation in government data strategy

The Open Data Institute (ODI) has published a report on data literacy and government approaches to data that criticises an over-emphasis on technical data analysis skills at the expense of understanding data and communicating about it in government and society.

The report’s authors say: “For all the apparently good work going on in government, there remains greater focus on the technical skills of [what the organisation calls] the ODI Data Skills Framework than the more contextual literacy side.”

The ODI Skills Framework is described elsewhere by the organisation as a means to “break down the complex landscape of data skills into the sets of skills required by different people in an organisation. It illustrates how technical data skills must be balanced with skills that enable data innovation”.

In appearance, the ODI’s framework is a colour-coded diagram made up of hexagons. On its left-hand side, it lists activities such as “innovating with data”, “working ethically” and “classifying data”. On its right-hand side, it lists activities such as “applying statistics”, “using data analytics” and “visualising data”.

The 48-page report, Data literacy and the UK government, was compiled by ODI staff Dave Tarrant, Milly Zameta and Violeta Mezeklieva, with Gavin Freegard, a special adviser to the ODI, as lead author of the paper.

It says it aims to “map the UK government’s activity on ‘data literacy’, as part of the ODI’s work supporting the development of the UK’s data economy”.

According to a statement that provides background to the report, the organisation defines data literacy as “the ability to think critically about data in different contexts and examine the impact of different approaches when collecting, using and sharing data and information”. It also contends that the concept “goes beyond the technical skills involved in working with data”.

The report’s summary mentions how “government approaches ‘data literacy’ for its own workforce, in its attempts to increase the use of data in policymaking and service delivery”, as well as how it seeks “to unlock the value of data across the economy”. It also registers both that “what the government does internally can also affect the wider economy and society, particularly given its role in leading by example, providing support (eg in publishing resources), and collaborating with others”, and that the public sector itself employs more than one-sixth of all workers in the economy.

One of its top-level findings is that the government has no guiding definition of data literacy, often eliding it into specialist, technical data skills. The report also highlights as a problem the fragmentation and duplication of responsibility for data across multiple bodies in and around government.

The report welcomes an improvement in data capability within the civil service, citing the Government Skills and Curriculum Unit (GSCU), the production of training resources by the Analysis Function and Government Statistical Service, and data masterclasses run jointly by 10DS (the Number 10 Data Science team) and the Office for National Statistics (ONS) Data Science Campus.

But it sees a disconnect between the government’s internal initiatives and its public-facing data literacy work. In particular, it deplores the fact that “most initiatives focus on the individual as employee, with an emphasis on the benefits to the workforce, rather on the individual as citizen and the benefits to them within society – the ‘data-driven economy’ rather than the ‘data-rich lives’ mentioned in the NDS [National Data Strategy]”.

The report singles out for praise “the AI Roadmap, from the independent, advisory AI Council, which has a particularly interesting, holistic view of literacy for all sitting alongside specialists skills for some and greater diversity in data-focused professions”.

But it laments: “There is no consistent definition of ‘data literacy’ across government, which reflects debates in academia and elsewhere. There is some recognition that ‘literacy’ means more than just technical, analytical and engineering skills, and that everyone – not just data specialists – needs some degree of data literacy.”

Data consultant and author Caroline Carruthers, former chief data officer for Network Rail, echoed the ODI’s criticism of an over-concentration, in and by the government, of the technical aspect of data literacy.

“You need to be able to bring data to life first before you get into anything else,” she said. “The basics are not being able to use Excel or read a report, but how you understand what you are looking at and confidently share data. Many organisations think data literacy is about learning how to do something in a BI tool or a package. That is not what data literacy is about.”