What this Sources Policy covers
This Sources Policy explains how we select and qualify evidence for an independent product publication, not a vendor sales page.
A useful citation should support the claim beside it and apply to the product or outcome being discussed. We do not treat the number of references, a prestigious logo or a search position as a substitute for reading what a source actually establishes. The Review Methodology explains the broader editorial process.
Which sources support which kinds of claims?
Different questions need different evidence: a checkout can establish its displayed price, while a health-outcome claim needs relevant scientific evidence.
| Question | Preferred source | Boundary |
|---|---|---|
| What is sold and how is it used? | Current product specification, delivered instructions or clearly attributed seller document | Marketing does not independently validate the product’s benefit |
| What does it cost and what are the refund terms? | Named checkout, current purchased-item terms and processor support information | Date the observation; do not promise an untested outcome |
| Does it affect a health or cognitive outcome? | Applicable human trials, systematic reviews and authoritative guidance | Mechanism, ingredient or different-stimulus research is not finished-product proof |
| What do customers experience? | Attributable independent reports with sampling and incentive information | Anecdotes do not establish typical results or a clinical effect |
| Who operates or endorses it? | Verifiable identity, authorization or primary institutional record | A competitor’s biography or decorative seal is not credential verification |
How scientific relevance is assessed
We compare the studied preparation and protocol with the actual product, population and outcome before transferring a finding.
- Study design, comparator, sample, exclusions and follow-up: what was tested and against what?
- Preparation and exposure: do the actual audio parameters, timing, route or dose match?
- Outcome: EEG activity, attention-task performance and everyday functioning are not interchangeable.
- Bias, funding, competing interests and replication: what could affect interpretation?
Favorable, negative and inconclusive results should be assessed using the same questions. A plausible mechanism is not an outcome trial. Arithmetic cannot reveal an undisclosed formulation or audio protocol, establish equivalence or prove synergy.
If only an abstract was inspected, that limit should be stated. We should not imply a full bias or funding appraisal of a paper we could not read in full. A PubMed entry is an index record, not an independent endorsement.
How commercial documents are used
Seller and checkout materials are useful for identifying the offer and dated terms; they are not independent efficacy or safety evidence.
Public articles identify important commercial documents, checked dates and relevant facts in plain language. Marked offer links are kept separate from evidence citations. A purchase button does not replace a source for a scientific claim, and a seller testimonial does not become independent because another affiliate repeats it.
When documents conflict, the review should identify the actual product and applicable clause rather than merge incompatible conditions. For example, a return policy naming bottles should not silently govern a different digital audio item. Material uncertainty must remain visible near the affected decision.
What counts as usable customer feedback?
A report needs an attributable source and clear limits; fictional, illustrative or unverified “verified purchase” stories are not presented as genuine customer evidence.
Seller-selected testimonials, incentivized reviews and affiliate articles may provide research leads, but their selection and financial relationships matter. We do not invent names, portraits, star ratings, satisfaction percentages or review counts to complete a page.
A complaint should be tied to the correct product and issue, with its status distinguished from a verified outcome. A small search sample does not show how common an experience is. Not finding a complaint does not establish that no complaints exist.
How citations and links are presented
Independent evidence belongs near the claim it supports; internal navigation and commercial actions have different purposes.
Scientific and authoritative guidance links should let readers inspect the relevant evidence. A references section helps orientation but is not a replacement for source-near support. Internal links point to the detailed owner of a topic rather than duplicating its full analysis.
Commercial links are marked and disclosed under our Affiliate Disclosure. Compensation does not change whether a study is applicable or whether a commercial fact has been verified. Customer-support links, where supplied, are identified as help resources rather than purchase recommendations.
How sources and changing facts are kept current
A checked date describes an actual inspection, not a promise that the offer will remain unchanged.
- Recheck price, package contents, billing, refund terms, support routes and safety warnings when they change.
- Evaluate relevant new research, corrections and retractions; do not automatically favor recency over quality.
- Update affected summaries, detailed owners, metadata and structured data consistently.
- Keep unresolved conflicts and missing dependencies explicit rather than quietly replacing facts with assumptions.
Lawful source versions or snapshots may be retained privately where feasible so an analysis can be reproduced. They are not a license to republish protected content or distribute purchased files.
What search and performance data can—and cannot—show
Search results help discover questions and sources, but rank, popularity and plugin scores do not prove accuracy or predict future performance.
We distinguish observed results from Search Console averages and third-party estimates. Without a reliable dataset, priorities remain qualitative; we do not fabricate search volumes, ranking difficulty, revenue or conversion rates. Affiliate clicks are not confirmed sales.
A bounded search cannot prove that every possible trial, complaint, warning or authorized seller has been found. Articles should communicate those limits naturally and avoid turning missing evidence into an unsupported absolute.
How to report a source problem
Use the Contact page to identify the article, exact claim and a source that changes or contradicts it.
A useful correction explains what the source directly supports, the date or version and why it applies to the product. We may revise a conclusion when better evidence warrants it; a seller objection or an affiliate commission is not, by itself, evidence that a claim is correct.
Use Contact and corrections for editorial issues. Do not send passwords, verification codes, payment details or medical records. This publication cannot handle a seller refund or provide individualized medical advice.