There is a movement towards increasing access to data, the development of dynamic displays, and evidence informed policy. A key challenge is to make data comprehensible – by creating better displays and by increasing the overall level of statistical literacy (SL) in the adult population. Here, we argue that the only plausible mechanism for improving SL is to use the data sources themselves to educate users; we explore alternative approaches, challenges, and opportunities, with examples.
One approach is to provide links to statistics tutorials (e.g. http://nationalstrategies.standards.dcsf.gov.uk/node/169418). These seem (to us) unlikely to be much used. A second is to generate a general interest in interface design. The demise of swivel (www.swivel.com now inactive) can be attributed to the fact that interesting data displays populated with trivial and unordered data are unattractive. A third general approach is to build the acquisition of SL into the process of data exploration.
To develop SL, the data must be inherently interesting. Hans Rosling's approach provides a model of an excellent interface, a charismatic presenter, and data sets that relate to global changes that people actually want to understand. Some important aspects of this model are unlikely to be generally applicable – in particular, videos on statistical issues that go viral.
The paper presents a variety of interfaces, and explores some approaches to embedded SL education. SMART Centre interfaces provide interactive displays and provocative questions; naïve users have been shown to demonstrate good statistical insights.
We also explore the idea that a hierarchy of heuristics can be created for novices at different levels of statistical sophistication which lead to judgments that a professional statistician would agree with (see Wild, Pfannkuch, Regan and Horton (2011, in press)).
Keywords: Statistical literacy; Heuristics; Dynamic displays; Interaction
Biography: Jim Ridgway has a background in psychology, and is a professor in the School of Education at Durham University, where he directs the SMART Centre. The SMART Centre creates interactive displays to make data accessible, in an attempt to help users convert data into knowledge. His research focuses on understanding and developing statistical competence in naïve users.