Is GIS a Reproducible Science?

The GIS work I have done so far is most closely aligned with the “GIS as a tool” position outlined by Wright et al (1997). As an undergraduate geography major, I have mostly encountered GIS in the context of the social and environmental problems that geography aims to address - particularly issues of land use, development, and environmental justice. In these contexts it lies within the problem it is used to address, and its implementation and repercussions are discussed in relation to that problem. Some more advanced classes, including this one, have started to bridge the gap into the “toolmaking” aspect of GIS, but until now none have delved deeply into GIS as a science in itself.

Whether or not the tool use and tool making aspects of GIS qualify as scientific research, I believe it is important to treat GIS as a science because of its widespread use and social impact. If GIS is treated as an unbiased tool that simply helps find an answer to some greater question and its methods and implementation are overlooked by the fields that employ it, I think there arises a much greater risk that issues with its methods and implementation contribute to inequity. Geography and cartography have a long and troubled history of colonialism and racism, and I think it is important to consider GIS as a science partly so that its practitioners are prompted to investigate and address how this history shows up in GIS. In a sense this puts the study of GIS squarely within the purview of disciplines like sociology, anthropology, and critical race studies, since the ubiquitous application of GIS in daily life gives it a great deal of power over things like housing, environmental protection, and urban planning that these disciplines already study.

Open source GIS can certainly help make science more reproducible, but its contribution will likely vary depending on how and where it is applied. The most basic tenets of open source - that code and data should be publicly visible wherever possible - are crucial precursors to making any study reproducible. Applying them in isolation will not solve the crisis of reproducibility though, and this is where the more slippery, difficult to define elements of open science could play a significant role. By enabling and empowering more people to “do science”, opening and democratizing scientific research could help advance the quality and number of scientific studies that meet all four of the National Academies’ criteria (Understanding Reproducibility and Replicability, pg. 44) for reproducibility and replicability.

References: Wright, D. J., M. F. Goodchild, and J. D. Proctor. 1997. GIS: Tool or science? Demystifying the persistent ambiguity of GIS as “tool” versus “science.” Annals of the Association of American Geographers 87 (2):346–362. DOI: 10.1111/0004-5608.872057

National Academies of Sciences, Engineering, and Medicine. 2019. Reproducibility and Replicability in Science. Washington, D.C.: National Academies Press. DOI: 10.17226/25303

Main Page