Course Introduction
PAF 516 — Core Requirement · MS Program Evaluation & Data AnalyticsAnthony Howell, PhD • Arizona State University
This course applies data science skills to the analysis of urban communities and neighborhoods. Students work with U.S. Census Bureau data, geospatial tools in R, and visualization frameworks to operationalize theories of neighborhood quality and community change. Through a cumulative applied project, students construct a composite economic hardship index, map spatial patterns, detect statistically significant clusters, track change over time, and deliver an interactive policy dashboard to a stakeholder audience.
Learning Objectives
- Construct reliable composite indices from census data using measurement theory principles
- Apply variable classification and spatial scale principles to create professional, audience-appropriate choropleth maps
- Integrate tabular and spatial datasets using joins, buffers, and proximity operations in R
- Detect and interpret spatial autocorrelation using Moran's I and LISA statistics
- Analyze demographic and economic neighborhood change across multiple ACS survey years
- Communicate analytical findings to non-technical stakeholders through policy briefs and dashboards
- Deploy interactive Quarto dashboards to GitHub Pages for public access
Cumulative Lab Project
Analytical thread across all modulesThe Economic Hardship Index you build in Lab 1 is the foundation for every subsequent module.
Starting in Lab 1, you will build a composite index of economic hardship using data pulled directly from the U.S. Census Bureau. The index combines multiple dimensions of neighborhood disadvantage — e.g., poverty, unemployment, and household income — standardized and aggregated into a single score for each geographic unit. You will construct the index at three spatial scales (county, census tract, and census block group) to see how the picture of economic hardship changes depending on the level of geographic detail. This index becomes the analytical thread running through every lab in the course — each module takes the same measure and asks a progressively more sophisticated question about what it reveals.
| Module | Lab | Core Question |
|---|---|---|
| Module 1 | Lab 1 | Is the index reliable? Build the index, assess Cronbach’s alpha, map at three scales |
| Module 2 | Lab 2 | How do we communicate it? Classification schemes, bivariate mapping, interactive GPU maps |
| Module 3 | Lab 3 | What else is layered on top? Spatial joins, buffer analysis, external data integration, and the Uncertain Geographic Context Problem |
| Module 4 | Lab 4 | Does economic hardship cluster spatially? Moran’s I, LISA hot spot maps, spatial autocorrelation |
| Module 5 | Lab 5 | Is economic hardship getting better or worse? 2013 vs 2019 ACS temporal change with pooled standardization and MOE significance testing |
| Module 6 | Lab 6 | Is change spatially structured? LISA trajectories, space-time Moran’s I, and multi-point trend analysis |
| Final Project | Interactive Dashboard | Does index composition change the story? Expand the Economic Hardship Index, analyze spatial sensitivity, and deliver policy recommendations |
Module Study Guides
Conceptual readings & lecture notesEach module guide covers the theory, readings, and key concepts for the week.
Measurement & Indexing
- Urban Measurement
- Spatial Constructs
- Economic Hardship Index
Classification & Spatial Scale
- Spatial Variable Classification
- Spatial Scales
- Modifiable Areal Unit Problem
Spatial Data Integration
- Spatial Joins
- Uncertain Geographic Context Problem
- Spatial Equity
Spatial Autocorrelation
- Tobler’s First Law & Spatial Weights
- Global Moran’s I
- LISA Hot Spot & Cold Spot Analysis
Spatio-Temporal Change (Part 1)
- Temporal Dynamics & Spatial Unit Matching
- Margin-of-Error Significance Testing
- LISA Hot Spot Overlay on Change Maps
Spatio-Temporal Change (Part 2)
- LISA Trajectory Classification
- Space-Time Moran’s I on Change Scores
- Multi-Point Trend Analysis
Lab Assignments
Hands-on R coding exercisesClick Tutorial to learn the method, then Assignment to complete and submit your work.
Measurement & Indexing
Build a standardized Economic Hardship Index, assess reliability with Cronbach’s α, and spatial visualization at three geographic scales.
Classification & Spatial Scale
Variable dichotomization and classification, bivariate and interactive mapping, and the Intraclass Correlation Coefficient (ICC).
Spatial Data Integration
Spatial joins and overlays, distance-decay analysis, and multi-buffer analysis.
Spatial Autocorrelation
Spatial weights matrices, Global Moran’s I, LISA cluster classification, and hot spot mapping with spdep.
Spatio-Temporal Change (Part 1)
Compare 2013 vs 2019 ACS tract-level data using pooled standardization, MOE significance testing, LISA hot spot overlay, and diverging color change maps.
Spatio-Temporal Change (Part 2)
LISA trajectory classification, transition matrices, space-time Moran’s I on change scores, and multi-point trend analysis to identify persistent, emerging, and dissolving clusters.
Final Project
Index sensitivity analysis & interactive dashboardExpand the Economic Hardship Index, analyze how spatial patterns shift, and write policy recommendations.
Student Guide
Read before starting. Explains the assignment, the measurement sensitivity analysis framework, and step-by-step instructions.
Instructor Dashboard (Baseline)
3-variable baseline Economic Hardship Index — Maricopa County, AZ. Use this as your reference when comparing your expanded-index results.
Student Dashboard
Right-click the link below and choose Save Link As… to download the QMD source file. Uncomment 1–4 index components, render, compare results to the instructor dashboard, and write your analysis.