How do we maintain public trust in data innovation?
Our digital lives create significantly more data than previously, and it is now easier to store, link, analyse and visualise than it has ever been before.
The Royal Statistical Society has been in the data game for over 180 years. We are now clearly in a period of considerable public and political interest in data with considerable hype and expectations placed on, for example, ‘big data’ and ‘data science’. This has largely stemmed from technological change: our digital lives create significantly more data than previously, and it is now easier to store, link, analyse and visualise than it has ever been before. New kinds of analysis including ‘machine learning’ and the increasing use of algorithms, bring traditional statistical methods together with new ideas from computer science.
Although further data innovation for societal good may be possible, public trust is required for this. Care.data was a recent attempt by government to increase data sharing of health records. It was badly executed and lost public trust, and so has barely progressed. A survey by Ipsos MORI for the Royal Statistical Society suggests that there is a general 'data trust deficit', and that public support for sharing personal data depends very much on who it is being shared with - and for what reason.
Public support for sharing personal data depends very much on who it is being shared with - and for what reason.
One positive response to this issue of trust is an increasing interest in data ethics. The Cabinet Office has just launched a data ethics framework to guide work in government. The Royal Statistical Society convened senior leaders to discuss the ethics of big data and proposed the need for a new Council for Data Ethics, which the Government has now committed to taking forward.
Another response to the issue of trust is the rise of bodies that check what is trustworthy. The UK Statistics Authority is the regulator for the production of official statistics (including that of HESA) and awards high quality statistics the brand of being ‘National Statistics’. It has in recent years, taken a more muscular approach to regulation and has ‘de-designated’ various statistics including the Retail Prices Index, crime statistics and UK trade statistics. This means they can no longer call themselves trustworthy National Statistics.
We are seeing the rise of independent fact checking bodies such as Full Fact, and the rise of ‘data journalism’.
The UK Statistics Authority also enters into the public domain when official statistics are being misused in public debate. It has rapped David Cameron on the knuckles for mixing up deficit and debt; Ed Miliband for how he talked about job creation in London, and most recently the Vote Leave campaign for its claim that leaving the EU could save the UK £350m a week. Alongside the Statistics Authority we are seeing the rise of independent fact checking bodies such as Full Fact, and the rise of ‘data journalism’ where the media is becoming more number savvy.
Open data is not just about transparency. It can create public and economic value.
A different way of addressing the gap in trust is through open data. Open data is data which anyone can use, access and share. Government has increasingly been opening up a wide range of datasets (e.g. expenses, floods) as part of a transparency agenda so the public has access to the same numbers as politicians. But open data is not just about transparency. It can create public and economic value. The most well-known example is where the opening up of transport data in London has created apps which allow us to check when the next bus is coming, when previously such data was guarded by Transport for London.
With numbers more pervasive than ever in society, we need a data literate public.
A final issue is that of data literacy. There was a time when data was not so ubiquitous and was largely handled by professionals, so you could trust they had the skills to know what they were doing. Now even in jobs such as nursing and teaching, data handling is becoming a common part of many people’s work. With numbers more pervasive than ever in society, we need a data literate public. Government has created a new ‘core maths’ qualification for 16 year olds who don’t wish to pursue A level maths but want to maintain their numeracy and improve their statistics. It is a good qualification but unless there is stronger backing from universities and employers we may not get the take up we want. The Royal Statistical Society has been playing its own part through training journalists in basic statistics (including a free online course), and also trains elected politicians.
As society becomes increasingly pervaded by data, it is critical we continue to develop the structures and tools to ensure what is done is trustworthy, and maintains public confidence.