Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Data Methods Equity topic
No spam. Unsubscribe anytime.
Training module teaches root-cause analysis and accessible mapping for equity-focused water data
Summary
An online module led by DataMaid and workshop co-organizers covered root-cause analysis, mapping best practices (projections, color palettes, accessibility) and reproducible R workflows using CalEnviroScreen and SAFER data.
An online training session in a two-part series on analyzing data for racial equity focused on root-cause analysis and mapping methods for water-quality and equity questions.
Hannah, partner at DataMaid, and Anna, workshop co-presenter and organizer, led the session. Anna described the core technique as iterative questioning: “ask why a lot, at least 3 times, usually closer to 5 or 6 times,” and recommended bringing community partners into root-cause work so data are paired with lived experience.
Why it matters: Presenters said equity-focused analyses require more than code — they require attention to accessibility, documentation and community engagement. The instructors encouraged participants to plan additional time for accessible visualizations and to include complementary presentation forms (tables, narrative summaries) for audiences who cannot interpret maps.
Mapping and data-work takeaways: The workshop covered common projection choices for California (WGS84 EPSG:4326 for point data, California Albers EPSG:3310 for statewide analysis), the need to reproject layers before spatial joins, and mapping options including choropleths, point maps/dot-density maps, and proportional‑symbol (bubble) maps. Presenters recommended ColorBrewer and Viridis palettes for color-blind accessibility and noted that legends and alternative data summaries help reach broader audiences.
Reproducibility and tools: Participants followed along in R using sf and ggplot2, and instructors shared a GitHub repository with code, sample datasets (CalEnviroScreen pesticide layer, SAFER risk data) and an equity data handbook. Hannah demonstrated filtering tracts at the 75th percentile and overlaying failing-system points with alpha transparency to reveal clustering. The GitHub code was offered to help staff reproduce analyses and to create reusable functions for prioritizing candidate systems by county, population served, or demographic variables.
Audience Q&A and practice: Breakout groups ran a ten‑minute root-cause exercise on Lamont, then reported hypotheses about exposure pathways (drinking water, bathing) and causes (agricultural pesticide use, legacy contamination, well age, underinvestment in disadvantaged communities). Several attendees suggested next steps: cross-check pesticide layers with land-use data, run multivariable regressions including race and income, and validate findings through local monitoring and community interviews.
Ending: Presenters closed by emphasizing that analytic work is an iterative practice that benefits from peer review and community partnership. They provided links to the GitHub repository, the equity data handbook and a feedback survey for future modules.
