Explaining data

Louise Marston - 09.11.2012

Data is becoming more and more prominent in fields as diverse as political campaigns, healthcare, business metrics and social deprivation. How do we bridge the gap between the data-savvy and the data-shy?

It seems that data has been appearing everywhere this week. Monday saw a debate at the British Academy about the use of Big Data in social science, including an interesting discussion of the ‘Reading the Riots’ research that used social media records. On Wednesday morning, we learnt just how accurate Nate Silver and his data-driven election predictions turned out to be. This week also saw the publication of new research on the digital landscape by the Government Digital Service (GDS), and the revelation that they have been developing a data dashboard for the Prime Minister. GDS is the team within the Cabinet Office behind the new, simplified gov.uk website, and have been doing great work to make government websites easier to use.

I’ve also been trying to write a post about data for a distinctly non-tech savvy audience (including my family). It’s a field that is steeped in jargon, and it’s made me think about how we communicate the potential of data to a less technical audience. How can we adapt services so that the data-savvy can feel empowered, without the rest feeling excluded?

The research published this week on UK internet users had some interesting results. They found that 82% of UK adults access the internet regularly or occasionally, and 56% have a smartphone. What surprised me was how well the numbers hold up amongst the older and poorer demographics: 70% of those 55–64 are online, and 79% of the ‘D’ demographic classification. (You can access the full dataset here.) There are still big divisions between the digital haves and have-nots, though - one of the reasons behind the research.

Overall, I think the general population is better equipped than we expect to handle numbers. What we need to avoid are the biases that creep into any assessment of statistics (a trait that professionals are just as susceptible to, as I highlighted in my last blog).

We encounter numbers every day - bank statements, political polls, medical risks. What makes the difference is good analysis and good presentation of data. A couple of years ago, Wired had a go at converting an uninformative sheet of blood test results into meaningful and useful outputs.

There is much that can be done without looking under the surface as well. We can all use Google without needing to understand the algorithms behind it. We are better off creating data-driven answers that are better for users than trying to show off the clever science behind them.