Methodology

Evidence quality is part of the answer.

Our framework is designed to show what an observation supports, what it does not support, and what a reader should verify next.

1. Define scope

We identify the marketplace, time, product interpretation, and decision question. A narrowly defined question reduces the temptation to generalize beyond the evidence.

2. Establish comparability

We examine model, variation, condition, quantity, seller, fulfillment, and included features before combining observations. A larger dataset is not automatically better if it combines unlike offers.

3. Preserve provenance and time

Material facts need a source context and observation time. Marketplace information is perishable; an old price or availability statement should not be presented as current.

4. Keep unknowns explicit

Missing shipping, seller, delivery, availability, or demand evidence remains unknown. It is not converted into free shipping, reliable fulfillment, availability, or zero demand.

5. Separate evidence quality from confidence

Completeness and reliability describe the evidence. Confidence describes certainty in an interpretation. Strong wording cannot repair incomplete evidence.

6. Consider downside and verification

We identify return, fragility, condition, delivery, price movement, and policy concerns. The final step is a verification list, not a promise of an outcome.