Qualitative Analysis

There are two definitions of qualitative analysis associated with GIS. One is the more scientific, meaning the use of noncontinuous data to categorize, explain, or predict geographical features. The other relates to a set of nonnumerical methods used to understand and to question how such features are conceived and represented in GIS. The first defines a way that GIS is practiced. The other is more concerned with specific GIS methods. This entry focuses on the latter of these two definitions but begins with a brief consideration of the former.

The more scientific meaning refers to categorical data analysis, where the inputs to or the results of analysis are either nominal (e.g., risk of noise pollution versus risk of light pollution, but neither obviously worse than the other) or ordinal (“rankable,” e.g., where severe noise is worse than mild noise but the actual increase in risk is not quantified). Examples include (a) layering a land use map (which has names to describe the class of activity occurring at locations) over other nonnumeric data, creating “qualitative” indicators of land degradation or environmental risk, and also (b) reclassifying a raster layer of continuous elevation data into a binary coverage (locations are either above or are not above a threshold height), overlaying it with other binary rasters (e.g., presence of soil contaminants: yes or no), and using a “sieving” (map algebra) procedure to identify locations that meet all criteria—designating those locations qualitatively as “suitable for development.” Here, “the qualitative” is some characteristic of geographical features that can be incorporated within an organizing frame-work, an ontology, to define and classify those features. Categorical data analysis is not necessarily numeric, but is underpinned by processes of enumeration, categorization, and counting that are more often regarded as quantitative, not qualitative.

When many in the social sciences and humanities talk of qualitative analysis, they mean to contrast it with the quantitative, eschewing statistical theory, mathematical modeling, and measurement in favor of observation, interviews, focus groups, critical reflection, textual analysis, and the interpretation of cultural artifacts within the context of their production. The focus is on how individuals or groups make sense of and ascribe meaning to their lived experiences and to the world around them; on understanding the values, beliefs, and motivations that lead people to act and to behave as they do in various socioeconomic and spatial settings; and on questioning the ontological assumptions that lead geographical features to be classified in the ways they are. If GIS have to do with computers, databases, and numbers, then how can this second meaning of qualitative analysis be relevant to geographical information science?

To begin an answer, it is helpful to distinguish between using GIS to generate qualitative data (in the categorical sense) and using GIS to display, analyze, and spatially join qualitative data (in the other sense). In regard to the latter, desktop GIS for some time have permitted nonnumeric information, including sounds and photographs, to be hyperlinked to positions on a map (so clicking on a GIS layer at certain locations causes particular images or noises to be produced). Newer, mobile technologies can be used with GIS to [Page 356]allow, for example, children to “tag” urban locations with “digital graffiti”—sounds and visions that are meaningful to them and that help researchers to better understand children's sociospatial practices, including the meaning they give to spaces that might appear as “meaningless” or empty on the standard cartographic products used by city planners.

...