Our methodology is the product.

Disclosure standard

Every study publishes the AAPOR disclosure set: sample source and type, field dates, completes by frame, weighting, quality screens, AI-use with human oversight, limitations.

Sample and quality

Verified sample vendors (Prolific-class verified supply, Cint, Dynata). Removal rates by screen reported per study — dedupe, speed, geo, trap, straight-line, LLM-detected.

AI and human review

Four human review gates: instrument design, data-quality disposition, interpretation, and narrative sign-off. AI never signs the story.

Compliance

CCPA/CPRA notice and deletion, GDPR Art. 89 minimization, service-provider contracts, AI-use disclosure. Deletion reaches AI-training artifacts.

Answers to the ESOMAR 37 (online samples) and 20 (AI services) questionnaires are published here.