Assessment

A trending detox tea removes toxins and aids long-term weight loss.

Evidence strength (0–100) = the quality-weighted share of evidence that supports the claim, out of all evidence for and against. Each source is weighted by study grade (RCT > cohort > case > preclinical > opinion) and, more gently, by sample size; predatory sources are cut 90%. Bands: ≥66 well supported, 40–65 mixed, <40 not supported. No meaningful evidence → “insufficient”, never a verdict from noise.

Evidence strengthNot supported · 3/100

The strongest supporting evidence is a opinion / anecdote. The strongest contradicting evidence is a randomized trial (n=1,200). Strength is the quality-weighted balance of supporting vs contradicting evidence; study quality outweighs sample size.

Strongest counter-evidence (always shown)

The honesty rule: we pin the single strongest study that disagrees with the claim — even when the claim is well supported — so the best counter-argument is never hidden from you.

Randomized trial: no lasting weight change versus placebo at 12 months.

Randomized trial · n=1,200

Supporting (2)

opinionOpinion / anecdote

Influencer testimonials and before/after photos.

Source: Social testimonials

case · n=18Case seriespredatory · low weight

Small case series in a non-peer-reviewed journal claiming rapid weight loss.

Source: Predatory journal paper

Contradicting (2)

rct · n=1,200Randomized trial

Randomized trial: no lasting weight change versus placebo at 12 months.

Source: Placebo-controlled RCT

cohort · n=3,000Cohort study

No toxin-clearance signal; effect consistent with a mild laxative.

Source: Metabolic cohort

See all sources, graded Get shareable card →

Compare with another assessed claim

verify.log

$ assess claim cmr1fimme0008atxfk588ejir
evidence gathered: 4 item(s)
support weight=0.81  contradict weight=22.77
strength = support / (support + contradict) = 3/100
verdict: Not supported
honesty rule: strongest counter shown

Ask AI

AI can be wrong. Not medical, legal, or financial advice.