{
  "study_id": "NCT07407595",
  "agent": "baseline",
  "run": 4,
  "p_primary_hypothesis_supported": 0.33,
  "effect_direction": "favours control",
  "effect_size_90pct_interval": {"low": -0.95, "high": 0.55, "metric": "Cohen's d for between-group difference in change in corticospinal excitability from baseline to post-cream (positive = excitability increases more / falls less in pain-expectation group)"},
  "moderator_prediction": "Higher pain catastrophizing / kinesiophobia participants may show larger excitability change (stronger threat appraisal), but with ~22 per arm this moderation will not be reliably detectable.",
  "claim_ids_used": [],
  "rationale": "Verbal-suggestion nocebo reliably moves self-reported pain, but expectation effects on objective outcomes are much smaller (OLP objective ~0.09). TMS I/O curves are noisy; with 22 per arm, 80% power requires d ~0.85. Pre-registered lab studies succeed ~44%; objective outcome lowers this. Flexibility (three Boltzmann parameters, two post time points, either direction) raises the chance of some significant group x time interaction. Prior experimental pain-threat literature leans toward reduced MEP amplitude in hand muscles during anticipation, so slight lean to negative d. Unmasked design and no actual nociception keep the expected effect modest."
}