From Data to Decision: Non-map-based Visualization of Geospatial Data Sets

The most common means of visualizing data sets with geospatial data is by employing map-based visualization models such as graduated symbol and choropleth. I argue that these models are in fact not the most efficient and effective visualization models for processing geospatial data, especially when the data set holds a notable quantity of location data. To support my argument I designed and developed two alternate models, one that did not use a map, and one that used an abstract version of one: a scatterplot model with geographic references, and a hexagon model. User tests were then performed to evaluate these models. This thesis describes the process, outcomes and future directions of my research. It also provides a literature review that addresses the definition and attributes of data visualization, a taxonomy of data visualization models, and a description of the mechanics of the visual cognition and colour theory employed