Introduction to Spatial Analysis
Jonathan Phillips, University of São Paulo
Political interactions are often shaped by the geographical context in which they take place; by the shape of electoral boundaries, by the segregation of urban neighbourhoods, or by the distribution of natural resources. This course provides an introduction to thinking spatially.
By the end of the course students will be able to reframe any political or social science question as a question of political geography: Why are some countries democracies? Which people voted for the current government? Why is politics becoming more polarized? How long can I expect to live? Students will also have the tools to organize and create geographic datasets which can address these questions and interpret and produce maps displaying complex geographic relationships. Finally, they will be able to implement spatial statistics used in a wide range of academic studies and policy analyses, including measures of spatial clustering (for example, of violence), segregation (by ethnicity), discontinuity (along borders) and manipulation (in gerrymandered elections).
Throughout the course we will discuss a range of applied examples from the rapidly growing fields of political geography and spatial analysis. Half of each day will be spent discussing conceptual and statistical issues in the use of spatial data, and half in the lab practicing and experimenting with datasets. We will explore datasets from published papers, and also from the wealth of spatial data published in Brazil, including both vector and raster data formats. The labs will use open-source software to ensure students can continue learning on their own beyond the course, including QGIS as a Geographic Information System and Stata/R for spatial statistics.
This course takes place between 14th – 18th January 2019.
TEACHING FELLOW Tainá Pacheco, Fundação Getulio Vargas
Day 1: Introduction - We start the course by asking why spatial and geographic relationships matter for politics? Drawing on recent examples from the literature and policy debates, our goal is to create a typology of geographic relationships to guide the datasets and research questions we will use in the rest of the course.
Day 2: Spatial Datasets – We discuss how to access, organize and represent data spatially, including how to use interactive Geographic Information Systems. We will learn how to merge geographic shapefiles with non-spatial datasets, project data and perform basic spatial calculations and transformations.
Day 3: Visualizing Spatial Data – Continuing our analysis of spatial data, we discuss more advanced topics including the analysis of point patterns and raster data, including an introduction to processing satellite images. We will also discuss techniques of map visualization.
Day 4: Measuring Spatial Relationships – To characterize spatial interdependence and patterns, we derive basic statistical measures such as the Moran’s I measure of clustering and learn how to calculate these ourselves using standard statistical software.
Day 5: Applications to Political Geography – Drawing inspiration from recent political science papers, we replicate core spatial analyses of clustering, segregation, boundaries and gerrymandering.