Spatial Data Operations
Content for Wednesday, September 24, 2025
Last class you learned a bit about joining attributes across tabular and spatial data. As you hopefully have learned, simply getting the attributes lined up doesn’t avoid the potential fallacies involved in spatial analysis. Doing that requires not just aligning who the data belongs to (i.e., via the attributes), but also ensuring that the scales of measurement align with our questions. Today, we’ll learn some of the spatial operations available in R that help us change the geometry of our data to better align the scales of measurement with the scales of our question.
Readings
Bigger Picture
Ecological Correlations and the Behavior of Individuals by (Lovelace et al. 2019) by (Robinson 1950) is the foundational paper on the ecological fallacy. Note that the language here is pretty out-of-date, but many of the issues and challenges he raises remain important 70 years later.
Incorporating geography into a new generalized theoretical and statistical framework addressing the modifiable areal unit problem by (Tuson et al. 2019) provides some really nice examples of how the MAUP can manifest in your data and why it matters.
Evaluating data stability in aggregation structures across spatial scales: revisiting the modifiable areal unit problem by (Nelson and Brewer 2017) introduces statistical techniques for assessing the stability of analytic results across different levels of spatial aggregation.
Technical Details
The introductory vignette for the
sfpackage has a lot of useful info onsfobjects and conventions.Section 2.2 on Vector Data and Sections 5.1-5.3 on Geographic Operations in Lovelace et al. (Lovelace et al. 2019) - for more details about vectors and geometric operations on vectors.
Section 3.1 and 3.2 of Spatial Data Science, a
bookdownproject by Edzer Pebesma and Roger Bivand (of thesf,sp,rgeos, andrgdalpackages).
Objectives
By the end of today, you should be able to:
Describe the Modifiable Areal Unit Problem (MAUP).
Link the inferential problems of the ecological fallacy to the statistical problems of the MAUP.
Modify the geometries of spatial objects to explore the MAUP.
Slides
The slides for today’s lesson are available online as an HTML file. Use the buttons below to open the slides either as an interactive website or as a static PDF (for printing or storing for later). You can also click in the slides below and navigate through them with your left and right arrow keys.