Skip to content

Connect Score v1.0.1 Validation Report

Connect Score distribution

Executive summary

  • Headline finding (Layer 3, criterion validity): Connect Score predicts where Louisiana invested $1.4B in BEAD funding (r = -0.499, 95% CI [-0.663, -0.288], q = 6.8e-05 after BH FDR correction; n = 64). Effect size: medium.
  • Convergent validity (Layer 2): Score correlates as theory predicts with educational attainment (r = 0.471) and rurality (r = -0.532). Poverty signal is r = -0.489. Poverty retains an independent signal beyond rurality (see colinearity diagnostic).
  • Methodology: Connect Score is a formative composite of three constituent dimensions (Availability, Affordability, Adoption). Component dimensions are appropriately distinct (max pairwise r = 0.305, well below the 0.85 collinearity ceiling).

Methodology framing

Connect Score is a formative composite index, not a reflective psychometric scale. Constituent dimensions are aggregated to form the digital-equity construct rather than to reflect a single latent trait (Diamantopoulos & Winklhofer 2001). This means components are by design distinct dimensions of digital equity, not items expected to inter-correlate. Standard reliability statistics for reflective scales (Cronbach's alpha) are not the appropriate validity test; we report it for completeness but rely on indicator collinearity (Layer 1 max pairwise correlation), convergent validity (Layer 2), and criterion validity (Layer 3) to establish trustworthiness.

Layer 1, indicator collinearity

The relevant Layer 1 test for a formative composite: do the three components measure distinct things?

  • Max pairwise correlation: r = 0.305 (threshold ≤ 0.85, PASS, effect size medium). Components are not redundant.
  • Pairwise: avail/afford = 0.281, avail/adopt = 0.305, afford/adopt = -0.225.
  • Reported for completeness, Cronbach's alpha = 0.335. Not applicable for formative indices; do not interpret as a reliability failure.

Layer 2, convergent validity

Connect Score should correlate as theory predicts with three independent signals (none used as score inputs). For each test we report Pearson r, Fisher-z 95% confidence interval, unadjusted p, BH FDR-adjusted q across the five inferential tests in Layers 2 and 3 (poverty, education, rurality, BEAD, unserved-unfunded), Cohen's effect-size label, and direction match against prediction.

  • Poverty rate (predicted negative): r = -0.489, 95% CI [-0.656, -0.276], p = 4.2e-05, q = 7.01e-05, n = 64. Effect size: medium. Direction matches prediction. PASS at the magnitude threshold |r| ≥ 0.4. Poverty scatter
  • Bachelor's-or-higher rate (predicted positive): r = 0.471, 95% CI [0.254, 0.642], p = 8.7e-05, q = 0.000109, n = 64. Effect size: medium. Direction matches prediction. PASS at the magnitude threshold |r| ≥ 0.4. Education scatter
  • USDA RUCC rurality (predicted negative): r = -0.532, 95% CI [-0.688, -0.329], p = 6.08e-06, q = 3.04e-05, n = 64. Effect size: large. Direction matches prediction. PASS at the magnitude threshold |r| ≥ 0.4. Rurality scatter

Colinearity diagnostic (poverty vs rurality)

Poverty rate and rurality both correlate with Connect Score. The Pearson correlation between poverty and rurality themselves is r = 0.607 (95% CI [0.425, 0.742], n = 64). The partial correlation Connect Score x poverty | rurality is r = -0.246, indicating poverty's marginal signal beyond rurality is meaningful. Poverty retains an independent signal after controlling for rurality. Both factors contribute to the digital-equity outcome.

Layer 3, criterion validity (the headline)

Does the Connect Score predict where Louisiana actually invests?

  • BEAD investment intensity (primary): r = -0.499, 95% CI [-0.663, -0.288], p = 2.72e-05, q = 6.8e-05, n = 64. Effect size: medium. Below the strong threshold (r ≤ -0.5); reported transparently as a partial signal. BEAD intensity scatter
  • FCC unserved-unfunded density (secondary): r = 0.053, 95% CI [-0.196, 0.295], p = 0.679, q = 0.679, n = 64. Effect size: trivial. No signal (|r| < 0.1). Unserved-unfunded scatter

Honest finding

The BEAD-intensity test did not meet the strong-pass threshold (r ≤ -0.5). Observed r = -0.499. Possible interpretations: (1) the score may need re-examination; (2) BEAD allocation may include political or process factors the score does not capture; (3) the signal may be partial. We report this transparently and acknowledge it as a v1.0.1 limitation.

Limitations

  • Sample size: n = 64 parishes. Confidence intervals reflect this. Findings are state-specific and do not generalize beyond Louisiana without re-validation.
  • Methodology version: v1.0.1 is the initial public version. Subsequent versions (v1.1+) may revise weights, components, or band breakpoints based on stakeholder feedback.
  • BEAD vintage basis: BEAD-funded locations come from the proposal-era Exhibit C Fabric, while the total-BSL denominator uses the current 2025-12-31 BDC Fabric. Absolute intensities therefore mix Fabric vintages, but parish rank ordering (which the criterion correlation depends on) is preserved.
  • Affordability proxy: structural affordability uses ACS B19001 income brackets at the $30K boundary as a 2%-of-monthly-income proxy for typical $50/month plan affordability. This is a coarse proxy; finer-grained income data would tighten the signal.
  • Cronbach's alpha is not applicable: the value reported (0.335) is not interpretable as reliability for a formative composite index, per the methodology framing above.

Reproducibility

  • Methodology version: v1.0.1
  • Inputs vintage: acs=2018-2022, fcc_bdc=2025-12-31, lifeline=2025-Q4
  • Score-input file SHA256 hashes: see exports/connect-scores-v1.0.1.json input_hashes field
  • Validation-input file SHA256 hashes:
    • acs_b15003.json: c9bcb98783abb7b3e0ec0aaef230226f8e858e9bd0cf4a65a4318c6812f97fb2
    • acs_b17001.json: 2c2df9c58e3ce6200950f099d353deceedb9cab6f8e632010c19864cad2e304b
    • bdc_22_Cable_fixed_broadband_D25_09jun2026.csv: 7acbe32a15fc5f59d960916521b441105e0c2d60afbf8d98e3560dddd77165e2
    • bdc_22_Copper_fixed_broadband_D25_09jun2026.csv: b9f36f37642b10b2744cfd74760f3271f4c66ae6ee7c358afeeebb342f218bb8
    • bdc_22_FibertothePremises_fixed_broadband_D25_09jun2026.csv: ba21c2c90f7f10272ca13f0807a0cf15d62eb9cd8732bf30486c84f44f7895e9
    • bead_locations.csv: 564a7a21495e317fcde69634b6c2ee146fae6228123e1e6597a410bdc06824b2
    • rucc_2023.csv: bc65b7d4ff352c3bddf3e3d02bf42336cf86635e5c2087e5e875239f32e1de7f
    • unserved_unfunded.csv: f9a547713052251de2c060a77933fdb1938fe8c00c0970f65de45e2ab7c6c272
  • Code: git log --oneline in this repo
  • Methodology spec: docs/superpowers/specs/2026-04-25-connect-score-redesign-design.md in the consumer app

Connect Score methodology v1.0.1, published Jun 18, 2026. Validation report · Sensitivity analysis