Review Process

Last reviewed: 2026-07-01 ← Editorial Framework

Purpose

This document describes how WaterBalanceTools reviews and validates its calculators, datasets, and knowledge content.

Automated Review

The primary review mechanism is automated. The build pipeline executes validate-datasets.js, validate-entities.js, and validate-trust.js on every build. These scripts reject the build if any required metadata is missing, any value lacks a source, or any trust component is broken. No page reaches production without passing automated validation.

Dataset Review

Canonical datasets are reviewed when: (1) a new authoritative source is published that contradicts existing values, (2) a user submits a correction with supporting evidence, or (3) the scheduled annual review date is reached. Review dates are stored in each dataset's lastReviewed field.

Calculator Review

Calculators are reviewed when: (1) an upstream dataset they depend on is updated, (2) a formula change is made, or (3) a correction is submitted. Calculator trust panels display the last reviewed date. Any calculator with a last reviewed date more than 12 months in the past triggers a review flag.

Formula Review

Formulas are reviewed when new scientific literature modifies the accepted equation for a calculation. Formula versions are tracked in data/trust/formulas.json. Changes are logged in the revision history.

Knowledge Content Review

Academy articles and reference pages are reviewed annually. Review dates are embedded in the data source (data/academy.json, data/reference.json) and displayed on each page.

Scope

The review process applies to all calculators, canonical datasets, formulas, academy articles, glossary entries, and reference pages. It does not apply to programmatically generated navigation, sitemaps, or schema markup, which are derived automatically from the reviewed content.

Review Frequency

Annual scheduled review for all content. Immediate review triggered by external standard changes or correction submissions.

← Back to Editorial Framework