THE EVIDENCE GLOSSARY

Know what the evidence actually says.

A source can be useful without proving every part of a claim. These definitions explain the terms we use when assessing an online store, with practical examples and their limits.

StraightReview LLCUpdated
01 / THE EVIDENCE GLOSSARY

Observed information.

Information directly seen in the material examined. For example, a contact page may display an email address. That supports the observation that the address was published; it does not establish that anyone answered a message or that the business operates from a particular building.

Read an observation together with its source and the date it was checked.

02 / THE EVIDENCE GLOSSARY

Retailer-published information.

A statement made by the seller, such as a delivery estimate, return window or product description. Attribution matters: “the policy states” tells you where a claim came from. It is different from saying a service was experienced or a promise independently confirmed.

Compare policies with your checkout terms and destination. Our delivery guide explains how to handle conflicting estimates.

03 / THE EVIDENCE GLOSSARY

Publisher-supplied context.

Research, prior familiarity or judgement provided by StraightReview’s publisher. This can explain the basis of an opinion, but it is not automatically independent corroboration. Relevant connections should be disclosed alongside the assessment.

The Oliver & Blake transparency note describes the publisher’s prior familiarity and the limits of that review.

04 / THE EVIDENCE GLOSSARY

Customer reports.

Accounts submitted by people describing an experience. A report can identify a useful question about sizing, support or returns. Repeating that report does not mean we verified the order or established how often the experience occurs.

Our guide to reading customer reviews explains how to compare accounts without turning an individual experience into a universal conclusion.

05 / THE EVIDENCE GLOSSARY

Independent corroboration.

Evidence from a separate, relevant source that supports a specific claim. Two pages copying the same retailer description are not necessarily two independent confirmations. Ask what each source directly knows and whether they rely on the same original statement.

A supporting source should match the exact fact being assessed. A business registration, for example, would not by itself establish product quality or a successful delivery.

06 / THE EVIDENCE GLOSSARY

Test purchase.

An actual order placed and documented for the purpose of examining an experience. A useful account would identify what was ordered, the relevant dates, what happened and which parts of the process were tested.

A single completed delivery would not prove every future order will succeed. The current Oliver & Blake assessment contains no test purchase; its public-policy checks should be read on that basis.

07 / THE EVIDENCE GLOSSARY

Assessment dates and page updates.

An assessment date records when the described evaluation was made. A source-check date records when a particular source was examined. A page-updated date can reflect a substantive clarification without a new check of every retailer policy.

Those dates are distinct. A review is not a live monitoring service. Follow the source links to check current terms before ordering.

08 / THE EVIDENCE GLOSSARY

Publisher confidence score.

A number from 0 to 100 assigned by StraightReview’s publisher as an editorial opinion. Scores above 80 carry our “Very likely trustworthy” label. It is not an average of customer ratings or a measured probability that an order will succeed.

Read the score explanation, retailer-specific findings and research scope together.

09 / THE EVIDENCE GLOSSARY

Editorial verdict and unresolved questions.

A verdict is the publisher’s judgement about the evidence and context described. Positive, mixed and negative are editorial descriptions, not customer-average scores or a guarantee about a transaction.

“Unverified” means the assessment has not established a point. It does not automatically mean the point is false. Read the assessment method and the review’s limitations before deciding how much weight to give a conclusion.

Clear sources. Useful context. A considered view.