Editorial Policy

Accuracy before volume

We would rather publish fewer pages that are right than a large site that is unreliable. Where we do not know something, we say so instead of filling the gap with a confident-sounding figure. That is why several pages here tell you to confirm a number with your recruiter rather than quoting one as settled — and why our retake page currently publishes no attempt limit at all.

Source hierarchy

When sources disagree, we prefer them in this order:

1. The current operational source for the specific fact. For test structure, counts, timings and scheduling, that is the delivery provider’s published information.
2. Current official policy documents from the Department of the Air Force and the Air Force Personnel Center, for retesting, score validity and testing security.
3. Official preparation material, which we link to rather than reproduce.
4. Our own measurements and product data, clearly labelled as ours and never presented as official.

Third-party test-prep publishers, forums, flashcard sets and marketplace listings are not sources for factual claims about the exam. They may be discussed on our comparison page, but nothing on this site states a fact about the AFOQT on their authority.

Where an older official document conflicts with a newer one, both are recorded and the newer authoritative operational source governs the live product. The conflict itself is noted rather than quietly resolved. Who applies these rules, and on what footing the site operates, is set out on the about page.

The source register

Every consequential factual claim about the exam is recorded with the claim itself, the source, the date the source states it was last updated, the date we last checked it, who checked it, which pages depend on it, and what would force an early re-check. We currently track 3 such entries.

An automated check fails the build when an entry passes its review interval, so a stale figure becomes a deployment failure rather than something a candidate discovers. Structural facts about the test are reviewed on a longer cycle than policy claims, which change more often, and anything can be re-checked early on an event trigger — an official announcement, or a report from someone who has just tested.

2 entries are currently unverified. Any figure that would rest on them is withheld from the site until a named reviewer has read the specific edition, and the affected pages say so on the page rather than in a footnote.

Review intensity scales with consequence

We treat AFOQT preparation as high-impact career content. A wrong claim about scoring, retakes or eligibility can change what somebody does with a commissioning attempt, so those pages carry the heaviest requirements: a primary source, a named owner, a visible review date, and a re-check trigger.

Study advice — how to sequence subtests, which respond fastest to practice — is editorial judgement, and we label it as such rather than dressing it as fact. Where we make a recommendation we try to give the reasoning, so you can disagree with it on the merits.

The three categories are kept visually distinct on the page: verified official facts carry a source and a date, editorial recommendations are written as recommendations, and platform estimates are labelled estimates.

Question standards

Every practice question is original, written from the publicly described subtests. We do not copy, paraphrase or reconstruct secure or recalled items, and we do not accept material from anyone who claims to have them. Full detail is on the question methodology page, including the publication gate that currently blocks the entire bank.

Advertising never touches editorial

No advertiser reviews content before publication. No commercial relationship earns a mention, a ranking position or a link in editorial text. We do not accept payment to include or exclude a product from a comparison.

The test of whether that is real is our comparison page, which publishes criteria and then lists where this site fails them. A comparison that always concludes in the publisher’s favour is not a comparison. See also the advertising policy and affiliate disclosure.

Corrections

Material errors are corrected on the page and logged publicly with what was wrong, what changed and how it was found. The log is append-only: entries are never edited or deleted, including old and unflattering ones. A disputed answer key is suspended immediately and stops being served while it is investigated, rather than staying live during the argument.

A factual correction is propagated everywhere the claim appeared, not just where it was reported, and each entry records whether that propagation is complete. Our first entry documents four subtest timings we had wrong during development — see the corrections log.

Claims we will not make

• That we have real, official or recalled AFOQT questions.
• That we can calculate your AFOQT score, percentile or composite.
• That we are affiliated with, endorsed by or approved by the Air Force, the DoD, AFPC or Pearson VUE.
• That any study plan guarantees a score, a selection outcome or eligibility.
• That a named expert reviewed something they did not review.
• Any pass rate, user count, testimonial or review score, because we hold no such data.

These are enforced as text patterns checked against our published copy, so a prohibited phrase fails the build rather than relying on an editor spotting it.

AI assistance

AI tools assist with drafting and engineering. They do not decide answer keys, factual claims, credentials or citations. The full position, including why this matters more on a study site than elsewhere, is on the AI content policy page.

Depublication

A page that can no longer be supported is corrected, merged or removed rather than left standing. Where a page is removed we redirect it to the closest genuinely useful destination rather than to the homepage, and where a claim is withdrawn the withdrawal is logged.

If you believe something here is wrong, tell us. A report explaining the reasoning is the fastest route to getting a bad claim withdrawn.