🟨 Andrei Cherny

• Strong negative correlations with all racial and minority indicators:

  • HISP_PCT: –0.43 → significant

  • MIN_PCT: –0.47 → significant

  • HISP18_PCT: –0.42 → significant

  • MIN18_PCT: –0.47 → significant

• R² values range from 0.18 to 0.22 — modest explanatory power.

• Total Population 18+: correlation near zero (–0.01), not statistically significant.

🟡 Summary: Cherny underperformed in diverse precincts (especially Hispanic and minority-heavy areas). His support was concentrated in whiter precincts, with no meaningful pattern related to precinct size.

🟥 Amish Shah

• Strong positive correlations with:

  • HISP_PCT: +0.48

  • MIN_PCT: +0.44

  • HISP18_PCT: +0.48

  • MIN18_PCT: +0.44

• All statistically significant (p < 0.001).

• Total Population 18+: mild negative correlation (–0.09) — not statistically meaningful.

• R² values around 0.20–0.23

🟥 Summary: Shah was the strongest performer in diverse, minority, and Hispanic precincts. His support dropped slightly in more populous (denser) precincts, but not in a meaningful way. Clear demographic alignment with racially diverse communities.

🟧 Andrew Horne

• No significant correlation with race or ethnicity:

  • Correlations: all near zero or slightly negative (–0.06 to –0.13), not significant.

• Only notable pattern:

  • Total Population 18+: +0.22 correlation, positive regression coefficient

  • R² = 0.05 → weak, but statistically significant

🟧 Summary: Horne’s vote share was not strongly tied to race or ethnicity, but he did slightly better in more populated precincts — possibly reflecting strength in moderate, suburban areas with higher turnout.

🟦 Conor O’Callaghan

• Moderate negative correlations with racial diversity:

  • HISP_PCT: –0.43

  • MIN_PCT: –0.41

  • HISP18_PCT: –0.42

  • MIN18_PCT: –0.41

  • All statistically significant

• Total Population 18+: near-zero correlation (–0.03), not significant

🟦 Summary: Conor performed more strongly in precincts with lower racial diversity — not as strongly correlated as Cherny, but still inverse to diversity. No clear link to precinct size.

Marlene Woods

• Very weak correlations with all demographics:

  • Ranged from –0.04 to –0.15

  • None particularly strong or meaningful

• R² values low (< 0.03), even if some are statistically significant due to large sample

• Mild inverse correlation with MIN_PCT and HISP18_PCT

Summary: Marlene’s support was not closely tied to any one demographic group. She had slight underperformance in more diverse precincts, but overall her appeal seemed broad-based and demographically neutral.