Public Health Snapshot

Clean and reshape into a tidy DataFrame. Save to data/processed/places_measures.csv.

Step 3: Download Census ACS Socioeconomic Data

Using the Census ACS 5-Year API (https://api.census.gov/data/2022/acs/acs5), pull:

• Median household income (B19013_001E)
• Health insurance coverage (B27001 table)
• Educational attainment — bachelor's degree or higher (B15003 table)
• Poverty rate (B17001 table)

for all four jurisdictions plus Virginia statewide.

Save to data/processed/acs_socioeconomic.csv.

Step 4: Benchmark Comparisons

Compare each jurisdiction against:

• Virginia statewide averages (from ACS and PLACES state-level data)
• US national benchmarks (from PLACES national estimates)

Flag any jurisdiction where a measure is more than 2 percentage points above the state average. Frame these as opportunities for investment, not deficits.

Save comparison table to data/processed/benchmark_comparison.csv.

Step 5: Composite Health Index

Build a composite health index for each jurisdiction:

1. Select the six PLACES measures from Step 2
2. Apply min-max normalization across the four jurisdictions (0 = best, 1 = worst)
3. Calculate an unweighted average to produce a single composite score
4. Rank jurisdictions

Save to data/processed/composite_index.csv.

Step 6: Visualizations

Generate three main visualizations:

6a. Jurisdiction Comparison Bar Chart

Grouped bar chart showing all six PLACES measures side by side for each jurisdiction. Include Virginia and US benchmark lines. Save to figures/jurisdiction_comparison.png and figures/jurisdiction_comparison.html.

6b. Radar Chart

One radar chart overlaying all four jurisdictions across the six health measures. Normalize so that outward = worse health outcome. Save to figures/radar_health_profile.png.

6c. Health vs. Socioeconomic Scatter

Scatter plot with median household income on the x-axis and composite health index on the y-axis. Label each jurisdiction. Add a trend annotation if appropriate. Save to figures/health_vs_income_scatter.png.

Chart standards:

• Use person-first language in all labels and titles
• Label whether rates are age-adjusted or crude
• Include data source footnotes on every chart
• Use colorblind-friendly palettes

Step 7: Community Health Narrative

Write a 500-800 word summary to reports/community_health_narrative.md that includes:

• Overview of the Richmond metro health landscape
• Key findings for each jurisdiction
• Structural and systemic context for any observed disparities
• Opportunities for community health investment
• Data limitations section noting that PLACES uses model-based estimates

Writing standards:

• Person-first language throughout
• Opportunity framing, not deficit framing
• No causal claims from this cross-sectional ecological data
• Note ecological fallacy: area-level data do not describe individuals

Quality Checklist

Before finalizing, verify:

• All rates clearly labeled as age-adjusted or crude — never mixed in the same comparison
• Confidence intervals included where available from PLACES
• Virginia and US benchmarks present in comparison outputs
• Person-first language in every narrative, label, and comment
• Opportunity framing used — no deficit-based descriptions
• PLACES methodology noted as model-based small area estimates
• No causal language (use "associated with," not "caused by")
• Ecological fallacy disclaimer included in narrative
• All data sourced from public APIs — no proprietary data