Geography’s Reproducibility Crisis
Open Source GIS as a Solution to Geography’s Reproducibility Crisis
- To what extent does open source GIS help solve the problems of the reproducibility crisis for geography? How?
Increasing access to the internet, software, and data around the world has led to an abundance of research, jobs, and opportunities for development and analysis. In the academic world, systems of landing tenure and making a name for oneself have led to incentives for researchers to overstate importance of their results and increase risk of bias in their studies, either conciously or subconciously. These factors have led to a “reproducibility crisis” in many scientific disciplines, not excluding geography. Many in the community are working to pull science out of this crisis, doing meta “research on the research,” but also seeing Open Source methods as a future for more robust, modern science. Led by Jon Claerbout in the 1990s, his “reproducible research movement” argued for standards of both openly shared data and code (NASEM, 32).
Open Source methods are the future for geography and GIS, and should be embraced at all levels of the discipline, especially in teaching and undergraduate contexts. Research is more easily reproducible when the software used is accessible for everyone, without any cost, geographic, or documentation barriers. Importantly, community developed software can be scrutinized as you can “poke around under the hood” for a more presice understanding of the program, study, or troubleshooting (Rey, 198)
- Are there problems with reproducibility and replicability in geography that open source GIS cannot help solve?
Open Source does not come without its challenges, however. Because software is constantly evolving and not being released in clean stages, newer versions could pose problems with backwards compatibility in a workflow or study design. In terms of making the data used public, which is the other side of the Open Source coin, there are many situations in which privacy or legal reasons would prevent data from just being appended at the end of a study. One aspect mentioned in Rey’s “Show me the code: spatial analysis and open source” is the developer-centric nature of open source projects as a negative. He says this leads to a lack of accessibilty, in that only those with adequate programming skills can participate in the development of the software (Rey, 195). I don’t think this is a realistic criticism of any open source software or discipline, as there are often steep learning curves in any discipline, and learning the craft is just a prerequisite of being involved in development of the tools. In the meantime, software is easy enough to use “out of the box,” guided by documentation written by the developers.
NASEM. 2019. Reproducibility and Replicability in Science. Washington, D.C.: National —–Academies Press. DOI: http://dx.doi.org/10.17226/25303
Rey, S. J. 2009. Show me the code: Spatial analysis and open source. Journal of —–Geographical Systems 11 (2):191–207. http://dx.doi.org/10.1007/s10109-009-0086-8