Combining Usability Techniques to Design Geovisualization Tools for Epidemiology

Abstract

Designing usable geovisualization tools is an emerging problem in GIScience software development. We are often satisfied that a new method provides an innovative window on our data, but functionality alone is insufficient assurance that a tool is applicable to a problem in situ. As extensions of the static methods they evolved from, geovisualization tools are bound to enable new knowledge creation. We have yet to learn how to adapt techniques from interaction designers and usability experts toward our tools in order to maximize this ability. This is especially challenging because there is limited existing guidance for the design of usable geovisualization tools. Their design requires knowledge about the context of work within which they will be used, and should involve user input at all stages, as is the practice in any human-centered design effort. Toward that goal, we have employed a wide range of techniques in the design of ESTAT, an exploratory geovisualization toolkit for epidemiology. These techniques include; verbal protocol analysis, card-sorting, focus groups, and an in-depth case study. This paper reports the design process and evaluation results from our experience with the ESTAT toolkit.

User studies in cartography: opportunities for empirical research on interactive maps and visualizations

“…Regarding administration, UCD scholarship is beginning to coalesce around a core set of methods both for interactive cartography (e.g. Marsh & Haklay, 2010;Robinson et al, 2005;Roth et al, 2015) and related fields (e.g. Cairns & Cox, 2008;Carpendale, 2008;Sweeney, Maguire, & Shackel, 1993).…”

“…Finally, qualitative and mixed-method research is now a staple for user studies in interactive cartography -particularly in the requirements analysis stage of UCD studies (e.g. Marsh & Haklay, 2010;Robinson et al, 2005;Roth et al, 2015) -and continued adaptation of social science methods from geography and related fields is a key methodological need.…”

“…These problems persist despite the now numerous recommendations in cartography to follow a 'user-centered' approach to the development and evaluation of interactive maps and visualizations (e.g. Cartwright et al, 2001;Fuhrmann & Pike, 2005;Kveladze, Kraak, & Elzakker, 2013;Lloyd & Dykes, 2011;MacEachren & Kraak, 2001;Robinson, Chen, Lengerich, Meyer, & MacEachren, 2005;Roth, Ross, & MacEachren, 2015).…”

HazMatMapper: an online and interactive geographic visualization tool for exploring transnational flows of hazardous waste and environmental justice

“…To support a range of potential users while facilitating visual thinking, designers increasingly consider the usability of the interactive map (Robinson, Chen, Lengerich, Meyer, & MacEachren, 2005), with large-screen design conventions moving toward a clear entry point into a central map that spans as much of the screen as possible, and a separation of interface controls in the layout by those configuring the interface and those providing additional context for advanced exploration (Roth, 2015). HazMatMapper includes four primary interface components: (a) a central map, depicting sites as proportional symbols scaled by imported hazardous waste with trade connections depicted as flow lines, or states as color-shaded choropleth maps (Dent, 1999;Slocum, McMaster, Kessler, & Howard, 2009).…”

Using treemaps for variable selection in spatio-temporal visualisation

“…[1][2][3] Information visualisation and geovisualisation techniques are an increasingly important means of assisting in the process of finding patterns and structure in large multivariate spatiotemporal data sets. [4][5][6][7] One approach to this process involves comparing variable-constrained subsets of a data set and selecting interesting variable combinations for further inspection. For multivariate spatio-temporal data sets, this can amount to hundreds or thousands of possible subsets.…”