Claims come first, prose comes last
Every city hub and school profile is built from a research dossier, not from memory. Before drafting begins, the researcher assembles a source table — every source consulted, with its URL, publisher, type and accessed date — and extracts the facts the page will assert into atomic claims: one checkable statement per entry, with its scope written in (academic year, grade, city), because fees and policies expire. A draft must be auditable claim by claim without re-reading its sentences.
Claim-level sourcing
Each claim carries one primary source and a confidence level. High confidence means an official first-party document — a regulator page, a school fee schedule, an inspection report. Medium means a reputable secondary source, or first-party material that is undated or ambiguous. Low — a forum or blog report with single corroboration — must be hedged in prose or dropped. Contradictions resolve in favour of the most recent first-party source, and the losing source is noted. On the published page, the source is linked at the sentence, so you can verify anything that matters to your decision.
The five source categories, every time
A dossier is not complete until every applicable category has been consulted — even when the result is “found nothing”: the official regulator (KHDA in Dubai, ADEK in Abu Dhabi, Singapore’s MOE, Saudi frameworks); the school’s own site and dated fee schedule (if a school hides its fees, that fact itself gets reported); parent forums, used as texture and leads and always attributed as reported experience; the school’s official social channels, with staged marketing noted as such; and relocation press for the moving-family context.
The unverified list, published in spirit
Anything we cannot confirm goes on an explicit unverified list in the dossier, with the reason — “no current schedule published”, “forum reports conflict”, “no authoritative waitlist data”. That material never appears on the page as fact. Where it is genuinely useful, it appears honestly hedged — “parents report”, “we cannot verify this” — because an honest gap beats a confident guess.
How we label judgment
Analysis and opinion appear only in clearly marked sections — “Our read” on profiles — and always carry a confidence label: High, Medium, or Limited. Document-based research says so. A profile we have not visited says so. Uncertainty is printed, not smoothed over. Fee and admissions claims are re-verified each academic year; regulator ratings on every new inspection cycle.
What we never do
We never invent fees, dates, statistics or quotations. We never publish a first-person experience we did not actually gather. We never take payment for coverage, rankings or recommendations — sponsored placements are labeled and cannot alter what research says. And we never publish placeholders: a city or school appears only when its research has cleared these gates.
Corrections
When we get something wrong, we fix it and say so — and the dossier is corrected in the same pass as the page, so the error cannot quietly regenerate. Material changes are noted on the page. Read the corrections policy.