Editorial Policy · Last reviewed 2026-09-18

Methodology

This page states how World Battery Hub produces, verifies, corrects and updates its content and data. It exists so a reader or an AI can see, before trusting a number, exactly how that number was obtained — and what its confidence level is.

Research methodology

Content is synthesized from multiple independent sources — including Chinese and English primary material — rather than translated from a single source. The method runs in four steps:

1. Collect — gather primary material: manufacturer datasheets, standards documents, government and industry data.
2. Cross-check — reconcile at least two independent sources for any factual claim; where they agree, record the agreement.
3. Label — attach the source, the test standard and the confidence level to every number.
4. Explain — where sources disagree, state the disagreement and its likely cause instead of silently choosing one number.

Source hierarchy

Sources are ranked by independence and authority, in this order:

1. Primary — manufacturer documentation, standards bodies (SAE, EN, JIS, IEC, UL, UN), government and regulatory data.
2. Secondary — industry publications, trade organisations and technical reviews.
3. Community — forums and user reports, used only to discover real-world questions, never as evidence.

A claim is only asserted as fact when it can be traced to a primary or a corroborated secondary source. Anything below that is stated qualitatively, or marked as unknown.

Data verification

Database fields (battery-group dimensions, CCA ranges, chemistry parameters) are written only when a source supports them. Where a value is a typical range, a standard-defined limit or a manufacturer claim, the field is tagged with that value_type so a reader or an AI cannot mistake a representative range for a guarantee. A missing value is written as null — never estimated to fill a gap.

Manufacturer-claim policy

Manufacturer data (company identity, capacity, certifications) is recorded at the verification level the evidence supports — claimed until it is independently corroborated, then cross-verified, then independently verified. We do not upgrade a manufacturer's own claim to "verified" without independent evidence, and we state plainly when a certification has not been independently accredited here.

Standards verification

A rating is only meaningful with its standard and test condition. Where we state a CCA or capacity value, we state the governing standard (SAE J537, EN 50342, JIS D5301, DIN 43539, IEC 60095, etc.) and the test condition. We do not compare values across standards as if they were interchangeable — the temperature and cutoff differ, and the difference is stated, not hidden.

Conflict resolution

When sources conflict, we do not pick the convenient number. We report the conflict, list the competing values with their sources, and state the likely reason for the difference (different year, scope, geography, or test method). Where a single value must be given, we give the range and its confidence.

Market-data method

Market figures (prices per kWh, GWh production, market size) are especially easy to misread because they vary by year, scope, geography, and battery definition (cell vs pack, production vs shipment, revenue vs market value). Each market figure is tagged with its metric, value, year, scope, source, method and confidence. A figure without that frame is not published.

Corrections

Errors are corrected in place, and material corrections are logged. To report an error, use the contact route on the About page. A correction does not rewrite history silently — the change is recorded with the date, the old value and the new value.

Update policy

Pages and database entities carry a last-reviewed date. Database entities are revised on a schedule and record a revision history where the underlying data changes. Content is not left to drift — the review date is part of the page, and a stale page is marked as such rather than presented as current.

Editorial independence

The platform has a commercial section (the product catalogue) that is clearly separated from the editorial and data content. Editorial conclusions, buyer guidance and the stance blocks are the Research Desk's own analysis and are labelled as such — they are not paid placement, and a commercial relationship does not change a rating, a standard or a confidence level.

Last reviewed: 2026-09-18

Where this method is applied

• Battery testing sub-site — the standards-referenced method pages.
• Battery data sub-site — the battery-group entity database with value_type tagging.
• Manufacturer sub-site — the manufacturer directory with evidence-level verification.
• Market sub-site — market data with metric / year / scope / confidence framing.