Within-candidate jump from preceding independent poll is +6.7 pp (t = 5.2, n = 132)

🟢 For 132 candidates whose self-sponsored poll was preceded by an independent poll in the same race (median gap 10 days), the within-candidate jump in error is +6.70 pp (one-sample t = 5.21, 74% of jumps positive).

This is the cleanest test of the "candidate commissions when privately believing they're leading" alternative. The comparator is the same candidate in the same race within a median 10-day window; the only thing that changes between the two polls is who's paying. A 7-pp jump on a 10-day window can't be plausibly generated by genuine momentum.

Restricted to time gaps ≤ 14 days (n = 77): mean jump +6.43 pp, t = 3.96, 77% positive. Magnitude survives the tight-window cut.

Sources.

Own analysis: build/table/regressions.csv (placebo block); AN-003
Cross-refs: H1 headline-sponsor-bias; headline-sponsor-bias finding

Cited analyses

AN-003 green placebo

For 132 candidates whose self-sponsored poll was preceded by an independent poll in the same race (median gap 10 days), the within-candidate jump from preceding-independent to self-sponsored error is +6.70 pp (t=5.21). The most intuitive counter to "self-sponsor when leading".