Supporting creativity and appreciation of uncertainty in exploring geo-coded public health data.

Background And Objectives: We present a prototype visualisation tool, ADVISES (Adaptive Visualization for e-Science), designed to support epidemiologists and public health practitioners in exploring geo-coded datasets and generating spatial epidemiological hypotheses. The tool is designed to support creative thinking while providing the means for the user to evaluate the validity of the visualization in terms of statistical uncertainty. We present an overview of the application and the results of an evaluation exploring public health researchers' responses to maps as a new way of viewing familiar data, in particular the use of thematic maps with adjoining descriptive statistics and forest plots to support the generation and evaluation of new hypotheses.

Methods: A series of qualitative evaluations involved one experienced researcher asking 21 volunteers to interact with the system to perform a series of relatively complex, realistic map-building and exploration tasks, using a 'think aloud' protocol, followed by a semi-structured interview The volunteers were academic epidemiologists and UK National Health Service analysts.

Results: All users quickly and confidently created maps, and went on to spend substantial amounts of time exploring and interacting with system, generating hypotheses about their maps.

Conclusions: Our findings suggest that the tool is able to support creativity and statistical appreciation among public health professionals and epidemiologists building thematic maps. Software such as this, introduced appropriately, could increase the capability of existing personnel for generating public health intelligence.