Global Statistical and Geospatial Framework e-Learning Tools Published

For nearly a decade, we have been discussing geospatial and statistical data on this blog. Such topics are more relevant than ever before. Central to these discussions is the question: How can organizations learn how to create and maintain the data necessary to support their priorities and goals? One way is through online courses. I am pleased to report that one such organization has created a course aimed at enabling organizations to accomplish the above mission: The Pan American Institute of Geography and History. This institute, which dates back to 1928, supports the Global Statistical and Geospatial Framework (GSGF). The GSGF permits the production of standardized and integrated geospatially enabled statistical data to facilitate data-driven decision-making. The resulting data can then be integrated with other information to inform and facilitate evidence-based decision making to support sub-national, national, and global development priorities and agendas, such as the 2030 Agenda for Sustainable Development. To this end, GSGF e-Learning courses have been developed by the leaders of Pan American Institute in collaboration with an organization called We Love Learning.

These e-Learning courses have been made possible via funding by the Pan American Institute of Geography and History and the tremendous efforts and work of the Expert Group on the Integration of Statistical and Geospatial Information and their development of the GSGF, and is a result of the commitment, collaboration, and work demonstrated by the Member States of the region participating in this project. This e-Learning course presents information that will support countries and users in understanding the GSGF – its value, application, infrastructure, and implementation requirements.

In my judgment, the course is nicely laid out with graphics and “knowledge checks” for the course attendee. English and Spanish versions of the GSGF e-Learning courses are available. The courses are structured 5 principles, including fundamentals of geospatial infrastructure, geocoded unit record data in a data management environment, common geographies for the dissemination of data, statistical and geospatial interoperability, and accessible and usable geospatially enabled statistics. There are high level summaries and an excellent set of glossaries, and other supporting resources.

The GSGF.

Reflections on recent GeoEthics webinar discussions

The GeoEthics webinar series from the American Association of Geographers and the University of California Santa Barbara, with support from Esri shares many common themes with this Spatial Reserves book and blog. These include surveillance (including location privacy), and governance (including regulation, data ownership, open data, and open software). I encourage you to watch the archives, including ethical spatial analytics (Feb 2021), responsible use of spatial data (May 2021), and others, and to keep tabs on the page to possibly watch an upcoming webinar live. Webinars are free and open to anyone, and AAG membership is not required, but you need to register ahead of time to watch them live.

On the webinar focusing on ethical spatial analysis, Dr Rogerson discussed examples pointing to instances where spatial dependence may confound the results of statistical testing. These practices raise significant ethical issues for public policy. Dr Vadjunec touched on an issue that we raised in this Spatial Reserves essay: Potential harm from location-tagged data and crowdsourcing. She also touched on privacy, the quality of data provided by volunteers and citizen scientists, issues raised by ethnographic research on very small numbers of human subjects, and the broader issues of trusting representations of the world as big data that may or may not be truthful when compared with real conditions on the ground. Dr Alvarez and Dr Bennett discussed maps as social constructs, how remotely sensed images are processed, and other pertinent related topics. Dr Sieber, in her discussion about artificial intelligence, discussed doorbell cameras, facial recognition, and other topics that will only become more important as time passes, and for which communities, including law enforcement, will have to make some important decisions on how, when, and why to use these tools.

Dr Goodchild, who has for myself and I suspect for many of us been someone we’ve admired and followed for a long time, made comments that made me realize that while we have made great strides in documenting data, we still have a journey ahead of us if we truly want another person or organization to be able to use the data to address a problem with the same workflows and inputs we used, for their own area of the world, or for the same problem with different variables or at a different scale.  This is reproducibility. Documenting data is only one part of enabling reproducibility. A key way this can move forward more rapidly is making sure that software companies, including my own, Esri, even more fully document the methods and models that are used for each analytical tool in their toolboxes. This could someday go so far as to send a message to the software user when the user is running an analysis tool, such as the presence of spatial dependence or some other factor. The bottom line is that all stages where spatial data is being processed should be documented and replicable, and efforts need to be made to estimate the uncertainties that are introduced. Accomplishing this rigorously is a noble and difficult to achieve goal.

But part of the responsibility will always be with the data users: Some GIS software such as ArcGIS Pro provide the user with history of the geoprocessing that was done as part of a project (such as a .aprx file). How can we encourage data users to include this history when they share their results with others? How can we encourage software developers to improve tools that will make data sources and methods easily discernible and transparent?

–Joseph Kerski