{
  "study_id": "NCT07611422",
  "agent": "baseline",
  "run": 4,
  "p_primary_hypothesis_supported": 0.15,
  "effect_direction": "null",
  "effect_size_90pct_interval": {"low": -0.25, "high": 0.45, "metric": "Cohen's d for positive-information vs neutral-information on PSS-10 at 4 weeks (positive = less stress with positive information)"},
  "moderator_prediction": "Expectancy effect slightly larger in the cold-shower arms, because a salient, effortful, novel practice paired with a resilience rationale is more credible than the same rationale for ordinary warm showers; the interaction test will be underpowered and likely non-significant.",
  "claim_ids_used": [],
  "rationale": "A single verbal rationale delivered once is a weak expectation manipulation aimed at a chronic, month-recall self-report outcome (PSS-10) four weeks later in healthy, non-stressed volunteers; the primer puts such effects as small at best (verbal suggestion is large only for acute outcomes, OLP non-clinical ~0.29, general placebo ~0.23). A realistic d of ~0.1-0.2 with ~60 per expectancy level gives power around 10-20%. Snowball recruitment, self-administered showers and uncertain compliance add noise. Slight upward allowance for self-report, an expectancy x time interaction path, and flexible analysis in a small academic study, but the likeliest report is a non-significant expectancy effect."
}