Today I had the great pleasure of meeting Dr Mei-Po Kwan from the University of Illinois at Urbana-Champaign, whose extensive work consists of analysing data with GIS in a mixed-method research. GIS has long been used by geographers and planners as a valuable tool to spatialize quantitative information, but Kwan has revisited the technique by building tools that incorporate qualitative data as well. That is a growing need considering new lines of research that acknowledge that conventional GIS is not enough to capt people’s feelings and perceptions (such as fear or place attachment). While past research usually relied on pattern recognition algorithms, recent technologies such as sensors and personal GPS have allowed for collection and analysis of individual complex data.
How residents perceive and use space is now a central concern for researchers focusing on people’s well-being and sense of place. The method called geo-narrative combines spatial patterns, analytical data and also social/behavioural information to generate an interpretive mode of analysis of large complex datasets (i.e. based on lived experiences). One important tool for narrative analysis, for instance, is called 3D-VQGIS, which permits adding images, audio files and texts to maps. Within the interface, the researcher is also able to code those texts according to nodes, a functionality common in qualitative analysis software such as NVivo. Another method presented was Ecological Momentarily Assessment (EMAs) which consist of live surveys of user’s sentiments on pre-determined or randomly selected spaces and time by questioning them via their smartphones. Sketchmap and cognitive mapping can also be used to assess user’s perceptions through participatory activities – all forms of connecting social aspects to GIS.
Furthermore, an additional factor that adds complexity to geo-narrative models is their ability to incorporate time. Points and georeferenced locations as we find in GIS are a stable information, but in fact, moveable points and populations are increasingly important. As Kwan pointed out, most of the times the key question is not related to where people linger but how they move and where their trajectory is affected by space and sentiments (Image 1).
Image 1: “Space-time paths of individuals collected with GPS can provide more accurate assessment of their exposure to environmental risk factors (e.g., traffic-related air pollution, carcinogenic substances, etc.) when integrated with detailed data about the
spatial and temporal variations of these risk factors.” Image and legend pulled from: meipokwan.org
Another idea that can be revisited is that of geographic units of analysis, which have traditionally consisted of census tracts or neighborhoods. The problem with cartesian and state imposed boundaries is that they do not correspond to residents mobility and living habits, thus, such spatial and temporal uncertainties can lead to misleading results which Kwan names the Uncertain Geographic Context Problem. However, linking models to behavior and travel patterns provide researchers with a new form of establishing such geographies.
Finally, the challenge remains on how to conduct detailed and statistically significant geo-narrative projects. Meanwhile, it provides us with a comprehensive form of building a new theory based on events in space and time which can serve as a basis for predictive models. Detailed literature on Geo-narrative and geovisualization is available on meipokwan.org.