Visualisation - it is not the data, it is what you do with it

The research described is focused upon the supporting the widespread adoption of data analysis skills as a means of empowering citizen engagement with the ever growing amount of open data. This has the potential to enhance the quality of engagement and effective use of data by helping minimize wrongly computed conclusions from given data. The work is applied in two technology areas where unskilled engagement with data is common. The first is that of the spreadsheet as a tool by which many analyze, plan and develop basic models for a number of purposes ranging from domestic budgeting to business planning. The second is more specialized language by which large data resources are commonly managed - the standard query language (SQL) for databases. The paper justifies the selection of these two areas and goes on to describe the approach taken to helping reducing potential user errors when working with both. This involves supporting users through visualizing the structures by which data is processed and characteristics of the steps involved in processing. This can be contrasted with the more common approaches to data visualisation that focus more up the outcome of data once processed. The two visualisation approaches are illustrated and initial results from the evaluation are described. Experience with using and evaluating these approaches has informed our understanding of the approach to visualization design.