{
  "study_id": "NCT07133529",
  "agent": "baseline",
  "run": 5,
  "p_primary_hypothesis_supported": 0.2,
  "effect_direction": "favours intervention",
  "effect_size_90pct_interval": {"low": -0.2, "high": 0.45, "metric": "Cohen's d (within-subject, crossover), uncertainty minus certainty in buffet intake"},
  "moderator_prediction": "Stress eaters (high SSES) and those with larger subjective stress or cortisol rises under uncertainty should show more extra intake, since uncertainty is theorised to act as a mild stressor that increases intake only in stress-hyperphagic people; the interaction will probably be underpowered at n=45.",
  "claim_ids_used": [],
  "rationale": "A pre-registered single-lab experiment with a subtle manipulation: meal timing withheld versus disclosed, in fasted young women. The theory (scarcity or insurance hypothesis) predicts more intake, but a one-session instruction change is a weak manipulation. Ad libitum buffet intake varies a lot from session to session, and the two visits are about 30 days apart, spanning cycle phase, so within-subject noise is large. With n of about 45 (fewer after dropout), 80% power needs dz of about 0.43, and the realistic effect is about 0.1 to 0.2. The buffet opens after testing, so both conditions are equally hungry, which dilutes any effect. Base rate for pre-registered psychology experiments is about 44%; adjusted down for weak manipulation and power."
}