Research, review and corrections policy
The evidence standards for observations, calculations, assumptions, reviews and publication readiness.
The claim determines the research required.
A title promising state costs and revenue requires evidence for both. A wage-only adjustment of a common scenario cannot be promoted as a completed local business study. A national business guide can explain a format without claiming local quotes or observed sales.
State research methodology v1 was approved by the owner on September 6, 2026. Its exact document is preserved and checked against the approval record on every build. Each business-state study must follow that version. Method approval does not certify the earlier reference scenarios, and missing local evidence cannot be replaced with shared commercial constants. Read the approved standard and its publication requirements.
Keep three kinds of information separate.
| Type | Required treatment |
|---|---|
| Observation | Record original source, exact claim or field, geography, period, units and retrieval date. Preserve the observation separately from a planning adjustment. |
| Derived result | Identify source inputs and formula. A payroll calculated from wages and assumed hours is a derived payroll amount, not a sourced business payroll observation. |
| Assumption or proxy | Explain why it is used, its scope and limitations, and what evidence could replace it. Suppressed values remain suppressed in the source record; an explicit proxy is not relabeled as a local fact. |
Use evidence that matches the decision.
Prefer original official data, responsible agency material and scoped first-hand commercial evidence. Several pages repeating one upstream source are not independent confirmation. Preserve contradictory evidence and explain differences in activity, geography, period and units.
Opening payments, recurring costs, revenue, paid labor and required permissions need separate supporting work. Identical values across states are acceptable only where their evidence supports reuse. Inventing differences is as misleading as hiding shared assumptions.
Technical checks and content review are separate.
Technical checks cover calculations, records, routes, metadata and generated files. Content review must assess the actual claim, evidence fit, material omissions, decision usefulness and exact published version. Word count, heading count and number of pages do not prove research completeness.
Presenting a page as completed state research requires the approved methodology, supported material inputs, traceable claims, recorded content review and accountable editorial responsibility. The current reference pages are publicly available with their assumptions and evidence limits disclosed.
How software assists the work.
Software assists extraction, drafting, calculations and consistency checks. Automated checks are reported as automated. No invented investigator, expert endorsement, completed interview or unperformed human review may be added to a page.
Correct the claim and its consequences.
For site support, policy questions and research corrections, email [email protected]. Include the page URL and enough detail to identify the question or claim.
Check a reported issue against the exact published source and version. A material input change requires reviewing affected financial outputs, narrative, comparisons and downloads. Keep a record of what changed, why, its evidence and the version affected. Source observation dates remain distinct from editing dates.
Prepare a correction · Publisher and scope · Current calculation definitions
Financial information disclaimer
Published research and calculations support business planning and education. They are not personalized financial, investment, tax or legal advice, and they do not guarantee costs, revenue, profit or financing. Estimates depend on the stated format, location, source periods and assumptions. Check the requirements and commitments that apply to your circumstances.