Collection of Open Source GIS project work during Spring 2021
I remember watching the shooting at Sandy Hook Elementary School unfold in the fall of 2012. I was at a swim meet in New York that weekend, and my friends and I were watching the news in our hotel as we ate a meal before the next session. One of those friends lived in Newtown, CT, and knew many of the students and teachers at Sandy Hook. Watching those news broadcasts was my first real exposure to the world of tragedy and disaster, and even then, social media entities such as Facebook and Twitter played an active role in spreading news and information about the event. That was nearly 10 years ago, and the presence of these entities has only grown more and more pronounced during the moment and aftermath of disasters. In their paper “The limits of crisis data: analytical and ethical challenges of using social and mobile data to understand disasters” (2014), Crawford and Finn discuss the ontological, epistemological, and ethical aspects of using social media data to influence how communities interact with disaster events. There are, without a doubt, nearly limitless positive and genuinely helpful applications for this data in crisis situations, as the past 10 years have demonstrated; the right people can get to the right places and provide necessary help more rapidly, the physical event itself can be recorded more accurately and measured to aid future disaster-prevention models, and alerts can be spread more efficiently to try and help people avoid harm before it even reaches them.
However, as Crawford and Finn report, it’s not as simple as just using the data and producing results. There are serious implications to how specific groups carry out some fundamental disaster response operations, that include logistical issues like how emergency teams travel to the disaster site, to how the community defines “disaster” in the first place. This data represents not just strings and lists of digits and letters, but the full range of human emotion - collected from the most vulnerable and essentially human people in the world. Similarly to the reality-to-analysis filters that any geographic experiment must pass through, there is a new ethical filter introduced when using real-world crisis data. “Context is critical here: people’s privacy preferences depend on their circumstances, and their choices shift depending on their situation” (Crawford and Finn, 2014). How is a researcher studying a disaster that happened 20 years ago supposed to accurately and respectfully discern context?
What, then, is the proper way to navigate this delicate balance? Clearly the benefits that come about from the use and analysis of this data are universal and significant, and I do not think that this data should be made inaccessible to researchers over concerns of due respect. I do think, however, that that data is inherently different from digits and letters collected in controlled or other real-world scientific experiments; that is to say, the vulnerability and trauma associated with the social media information is just as much a part of that dataset as the geospatial coordinates themselves. Just as there is a way to “respect” the geospatial data by validly and accurately analyzing, and then illustrating it in accessible and useful ways, there are ways to respect the human essence of this data. Being absolutely certain no personal information is shared is the easiest way to respect this essence, but more holistic and appreciative ways could include acknowledging the struggle and loss of those who lived the disaster being studied, stepping out of a scientific tone for a sentence to point the audience towards an appropriate fundraiser, or utilizing one’s role as an expert in the field to influence real-world decisions (although, for this to happen beyond using the produced study would involve a level of community engagement well beyond the scope of the author’s current field of study).
Part of the open source science revolution is understanding and internalizing the multi- dimensional aspects of science and scientific analysis. In order for the science to be good and valid, it must be reproducible, and when it comes to studies utilizing disaster data that incorporates a range of human emotion, that aspect must be addressed and given due respect in order to meet this criteria. As the world experiences more and more natural disasters, and as social media pervades deeper and deeper into the fabric of human social interactions, conducting science at these crossroads will become more important and more high-stakes. Let’s get it right from the start.
Reading:
Crawford, K., and M. Finn. 2014. The limits of crisis data: analytical and ethical challenges of using social and mobile data to understand disasters. GeoJournal 80 (4):491–502. DOI:10.1007/s10708-014-9597-z
Submitted May 03, 2021