Spatial Analysis for Psychologists: How to Use Individual-Level Data for Research at the Geographically Aggregated Level

Psychologists have become increasingly interested in the geographical organization of psychological phenomena. Such studies typically seek to identify geographical variation in psychological characteristics and examine the causes and consequences of that variation. Geo-psychological research offers unique advantages, such as a wide variety of easily obtainable behavioral outcomes. However, studies at the geographically aggregate level also come with unique challenges that require psychologists to work with unfamiliar data formats, sources, measures, and statistical problems. The present article aims to present psychologists with a methodological roadmap that equips them with basic analytical techniques for geographical analysis. Across five sections, we provide a step-by-step tutorial and walk readers through a full geo-psychological research project. We provide guidance for (a) choosing an appropriate geographical level and aggregating individual data, (b) spatializing data and mapping geographical distributions, (c) creating and managing spatial weights matrices, (d) assessing geographical clustering and identifying distributional patterns, and (e) regressing spatial data using spatial regression models. Throughout the tutorial, we alternate between explanatory sections that feature in-depth background information and hands-on sections that use real data to demonstrate the practical implementation of each step in R. The full R code and all data used in this demonstration are available from the OSF project page accompanying this article. Translational Abstract Psychological characteristics are unequally distributed across space. In recent years, psychologist have become increasingly interested in revealing such geo-psychological differences and relating them to consequential real-world outcomes. These studies at the geographically aggregate level are important as they offer some unique advantages. For example, aggregate-level studies enable researchers to link psychological variables to behavioral outcomes that are difficult (or impossible) to collect at the individual level. However, studies at the geographically aggregate level also come with unique methodological challenges. Specifically, doing research with spatial data confronts psychologists with data formats (e.g., geographical shapefiles), measures (e.g., to quantify geographical clustering), and statistical problems (e.g., spatial autocorrelation) that are not part of their conventional training. To bridge the gap, this paper seeks to provide psychologists with a ready-to-use and cost-free toolkit to successfully use spatial methods in their work. To that end we provide a step-by-step tutorial that walks the reader through the entire process of a geo-psychological research project. This process spans from the spatial aggregation of individual-level data to producing informative maps and performing spatial regression models. To demonstrate how to implement each step we use real, open-access data and provide full analysis code in R.