Evidence Judgment
Notice what is supported, incomplete, biased, or still only an estimate.
Judgment for an AI-shaped world
AI can produce an answer.
Can it find your judgment?
ThinkProof reveals how you evaluate evidence, challenge assumptions, and make decisions when an AI recommendation sounds convincing—but the case is not complete.
15–20 minutes · Human review for uncertain results
Not a knowledge test.
Not a personality label.
A record of how you think.
What ThinkProof measures
Good judgment leaves evidence.
The assessment looks past the final choice. It examines what you noticed, what you questioned, and how responsibly you updated.
Assumption Detection
Surface the hidden claims that turn observations into recommendations.
Trust Calibration
Give AI evidence and human context the weight each one has earned.
Decision Under Uncertainty
Act proportionately without pretending that every unknown has disappeared.
Adaptability
Update when new evidence matters while carrying unresolved constraints forward.
The experience
Three decisions.
One visible reasoning trail.
You do not need specialist knowledge. You need to notice what matters and explain why.
01
Challenge
Work through three realistic decisions
Vendor selection, operational efficiency, and customer churn reveal how you reason when the evidence is mixed.
02
Review
Explain what changed your mind
New information tests whether you update proportionately—not whether you simply change position.
03
Profile
See the judgment behind your answers
Your profile connects conclusions to response evidence and routes uncertain results to human review.
Sample result
A profile you can inspect,
not just a score.
Each conclusion is tied to evidence from the participant’s own responses. Low-confidence or watchlist results remain unscored until a reviewer completes them.
✓
15 of 15 dimensions reviewed
Profile ready
ThinkProof profile
Sample Participant
01
Evidence Judgment
Strong
02
Assumption Detection
Strong
03
Trust Calibration
Calibrated
04
Decision Under Uncertainty
Strong
05
Adaptability
Strong
Signal summary
“Carries unresolved material constraints forward while updating proportionately when new evidence changes the case.”
Illustrative profile · Not a credential or hiring recommendation
Privacy by design
Your identity is not part of the AI scoring prompt.
After consent, the scoring process sends redacted answer text for analysis. Names and email addresses are excluded. Existing responses from before the consent cutoff are not sent for scoring.
- 01 Only consented response text is evaluated
- 02 Identity fields stay out of the scoring request
- 03 Low confidence and pass disagreement trigger human review
- 04 Review-gated scores stay blank until evidence is added
ThinkProof Challenge 001
Ready to see how
your judgment shows up?
The ThinkProof challenge is live. Work through three professional decisions and explain what evidence changed—or did not change—your position.
Begin Challenge 8–10 minutes · Review and accept the consent statement before submitting.