Card Sorting for Cartographic Research and Practice
Roth, Robert E., Finch, Benjamin G., Blanford, Justine I., Klippel, Alexander, Robinson, Anthony C., MacEachren, Alan M., Cartography and Geographic Information Science
Categorization is a fundamental way in which humans make sense of their world, and therefore is a critical way in which knowledge is organized (Abler et al. 1971; Lakoff 1987; MacEachren 1995; Margolis and Laurence 1999). The card sorting method is a knowledge elicitation technique designed to reveal the conceptual structures, or categorizations of targeted individuals (Cooke 1994). Card sorting requires participants to organize a set of instances, referred to as cards due to the original use of analog index cards into internally-homogenous groups, or categories according to similarity along an identified sorting principle or criterion (Spencer 2009). Depending on the card-sorting variant, the sorting criterion and categories may or may not be provided, and the kind of information placed on the cards may be different (text, images, icons, etc.).
The potential of card sorting for both cartographic research and practice is substantial. As the number of symbols included on the map grows, it becomes increasingly unlikely that map readers will be able to remember and thus immediately identify the feature type represented by each symbol. Organizing map symbols into broader categories improves the memorability of symbols by imposing a grouping rule, better delineates key themes within the map when the symbols are designed to reflect their higher-level category, and structures map legends for improved symbol reference. Card sorting is particularly promising for the design and organization of qualitative point symbol sets that use pictorial symbols to signify differences in kind, as the method works best for sorting items on the same semantic level (Osgood et al. 1957) into nominal categories (McGeorge and Rugg 1992) that have a single-level hierarchy (Wood and Wood 2008). Additional benefits to the application of card sorting to cartographic design include categorization of previously ungrouped map symbols, critique and refinement of existing symbol categorizations, and identification of missing or ambiguous symbols within a symbol set. Finally, the card sorting technique may be useful for integrating expert knowledge into a mapping system that automates the design of symbols and associated legend layouts.
The research reported here is part of a broader line of work at the Penn State GeoVISTA Center on map symbology for emergency management and first response (Akella 2009; Koua et al. 2009; Robinson et al. 2010). Specifically, this research is part of our work to develop a collaborative, yet distributed process for designing and sharing 'mission-specific' map symbol sets to support the range of mission areas for which government agencies use maps. Our focus in this article is on the practical application of card sorting to design a single symbol set for a specific mapping context, rather than its use as a scientific mode of inquiry to produce generalizable and repeatable results. The framework we present for application of card sorting in cartography, however, is designed to be relevant to support both research and practice.
This article proceeds with five additional sections. In the following section, we review relevant work on card sorting in the fields of psychology, usability engineering, and GIScience. We then outline a framework organizing variants of the card sorting method when applied to cartography In the fourth section, we introduce our case study and experimental design. As a demonstration of the applicability of card sorting as a method, and as initial input to the larger research project on map symbology cited above, we conducted a card sorting study with twenty Penn State undergraduates. Specifically, the study was designed to evaluate an existing emergency mapping qualitative point symbol standard designed by the Federal Geographic Data Committee (FGDC) Homeland Security Working Group (ANSI INCITS 415-2006) to support emergency management and first response. We present and discuss the results of the card sorting study in the fifth section and offer concluding remarks and future directions in the sixth and final section. …