Why geographic data science is not a science

Abstract

"Data Science" has taken many disciplines by storm. And for a good reason: New forms and unseen quantities of data enter nearly every scientific field, substantially changing the ways how scientists do science, and potentially allowing them to answer old questions or to pose them in novel ways. The recent success of Data Science is also reflected in corresponding study programs and curricula and the emergence of specialized branches, such as Geographic Data Science (GDS). Some researchers, therefore, claim that Data Science and GDS should be treated as autonomous scientific disciplines, while others fear that it sells nothing but old wine in new bottles. In an attempt to sober the discussion, we investigate GDS and Data Science from the perspective of meta-science. We provide arguments why today's GDS and Data Science should be seen as an interdisciplinary community of practice of data-driven scientists, rather than a scientific discipline. We also discuss what is missing for GDS and Data Science to become genuine scientific disciplines.

New steps against scientific gerrymandering: A new operational definition of geographic information science

“…Following Scheider et al, (2020), in our proposal real world information will be considered by empirical sciences (such as Demography, Biology, Geology, etc.) defined as opposed to formal sciences (such as Mathematics, Computer Science, Logic or Linguistics) that do not require experimental validation.…”

The Synergy of Data Science, GIS Spatial Analysis and Knowledge Management as a Path to Sustainability Insights

“…As side-effects of the overwhelming digitalization, the massive advancements in computation capabilities and information and communication technologies affected approximately all disciplines and professions. It was thus argued that data science had shifted and redesigned all scientific disciplines, by means of adjusting their adopted approaches [5].…”

The Right to Rule by Thumb: A Comment on Epistemology in “A Route Map for Successful Applications of Geographically‐Weighted Regression”

“…Sometimes, this work developed into an open debate about different versions of the SAR route map (Florax, Folmer, & Rey, 2003;Florax, Folmer, & Rey, 2006;Hendry, 2006). In general, debates about which routes were safest only came about when stable and performant implementations of spatial econometric models began to attract distinct communities of users outside of the initial innovators-an important step in defining a community of practice from which richer domains of science can emerge (Scheider, Nyamsuren, Kruiger, & Xu, 2020). This also happens to be true for Bell, Fairbrother, & Jones (2019), who seek to resolve disagreements about an important starting step in route maps of multilevel model terrains.…”