A public culture of statistical literacy

The premise of of wider statistical education for all is rooted in the idea of a societal change. More than the data skills needed in specialised roles in policy and management, or the obvious technical roles, citizens stand to benefit substantially. The potential of data to benefit people is expressed clearly in the report on Data Governance published by the British Academy and Royal Society:

Data can empower individuals and communities to exercise rights that they
previously found difficult to exercise effectively. (p.53)

At present there are campaigning organisations such as Full Fact and the Alliance for Useful Evidence, as well as specifically scientific organisations, engaged with public statistics. While these challenge inaccurate usage of statistics and unevidenced policy proposals, this is a bit different to ’empowering communities’. Indeed they could be described as scrutinising the minutes rather than setting the political agenda.

Post truth

The establishment has experienced a crisis which they have dubbed a transition to a ‘post truth’ politics. The reaction to this has involved many seminars and events bemoaning the success of politics which have not followed their rules. A much derided bus had a message which was factually flawed but whose rhetoric resonated with the demos. And the statements of Donald Trump has been designed to engage and emote, not persuade, like the level of hidden unemployment.

The elite media have expressed outrage matched with impotence even as some point out the difference between taking politicians literally rather than seriously. The books published by such as Evan Davis give a clear indication that they have not reflected on the experience, let alone learned from it. The set piece media events such as his Newsnight are aggressive confrontations which follow the political agenda.

The democratic agenda

For all the importance of the national economy, the deficit, growth etc, these only relate indirectly to people. This was well captured by an audience member contesting the ownership of GDP – the elite had not noticed this was not translating locally. In this sense, the new industrial strategy with its regional concerns and concerns for local growth was an improvement. And some media had taken up this challenge, reporting more from ‘Anywhere but Westminster‘, letting people contest the agenda.

However, much that is important to people remains too local to be reported nationally. And so it may not be discovered at all, until it is too late, as in the tragic case of Grenfell Tower. Even the threat of media scrutiny has now prompted resignations, but the documentation is clear enough that journalism might have changed actions at the time. It is no wonder that Private Eye reports record circulation as it is one of the only organs to maintain independence from the herd.

How can data help?

The Royal Statistical Society is among those who organisations which could be accused of taking too national a focus. But this week they published the policy basis to support their data manifesto position on local data. In part this has been provoked by the closure of the Neighbourhood Statistics site which provided local indicators at a postcode level. But there is a depth of thinking that the use of data is at the level of the concern, which is typically national only in aggregate, e.g. for health inequality.

The availability of data at sub-national level is less common than it should be but communities should also be able to develop their own solutions. Anticipated development of guidance on local evaluation of transport investments by the DfT is encouraging as an example. But making use of it requires considerable statistical skills both among analysts, and among leaders, to understand what data can show, and to let the data suggest and answer our questions.

Governance

People are at the heart of the whole enterprise, a point that the data governance report recognises. The proposed solution has variously been described as a council, commission or convention on data ethics, use and governance. But the principles underlying it, and the lack of a hierarchical authority or dictatorial model, are inspiring. While some processes are emerging in sectors like health which depend on data and have large administrations, others will need proportionate support.

Local organisations, both private and public, have data on the people they interact with, and should use this to shape their activity. Their capacity to establish rigorous procedures to govern this may be much less, but they can rely on more direct access to their public. And it is this continued interaction with people which is the proposed solution, so they understand the nature and purpose of data, how its use may benefit them, if they are content for it to be used.

A public culture

At present, there is a national and adversarial attitude to statistics in pubic life. Different groups bring their own evidence to bear, without clear resolution of how this pertains to the concerns of people. Some things are dismissed out of hand for minor technical concerns, and others accepted without challenge. And where there is concerted scrutiny, it can focus on a narrow expectation of veracity without considering how claims were interpreted by people.

Media reports about statistics are often the latest data in a long series, which users scrutinise for informative trends. They do not correspond to local levels of concern, and the categories they use often do not describe these concerns either. The public need to be confident in asking questions and challenging, not as Russell Brand did, dismissing Evan Davis’ attempt to present him with a graph, but contesting standard categories and the blanket authority claimed.

Empowerment

Data can empower, but only if people have the capacity to utilise it. Statistical literacy is truly a literacy, that people need to be able to produce their own accounts of what data shows. And they need to be critical readers of the presentations of others, being able to respond revise and adapt initial analyses. Present discourse is dysfunctional in that it excludes some and venerates others without appropriate scrutiny. The culture needs to change.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s