{
  "study_id": "NCT07731568",
  "agent": "framework",
  "run": 2,
  "p_primary_hypothesis_supported": 0.3,
  "effect_direction": "favours intervention",
  "effect_size_90pct_interval": {"low": -1.5, "high": 3.5, "metric": "Between-group difference in GAD-7 change, 0-21 scale, at 4 weeks (positive = larger reduction with NPT)"},
  "moderator_prediction": "Higher-baseline-anxiety students gain more from NPT over CBT: their anxiety is less movable by cognitive restructuring (argument) and responds to enacted, feared action and acceptance; low-baseline students show floor effects in both arms.",
  "claim_ids_used": ["PM-0056", "PM-0238", "PM-1353", "PM-0851", "PM-1345", "PM-0218", "PM-0081", "PM-0383", "PM-0939"],
  "rationale": "The framework favours NPT slightly: CBT's core cognitive restructuring is argument, a first-order deposit that barely moves belief (PM-0056, PM-0238), while NPT's 'do what you fear' and acceptance principles demand enactment, which installs (PM-1353, PM-0851, PM-1345). NPT's framing in traditional Chinese culture is compatible with participants' cultural programming, lowering the compatibility-check cost (PM-0218, PM-0081, PM-0939). But CBT also contains behavioural experiments, both arms get the same group delivery (PM-0383), both are online and 4 weeks, and n=55 gives power only for large differences. Sponsor allegiance raises the chance of a significant result, yet superiority over an active comparator in so small a trial remains less likely than not."
}