Editorial Standards
Every one of the 46 pages here is researched, written and reviewed by Christoph Ballmer. This page sets out the steps that takes, which of them are enforced automatically, and how to get an error fixed.
Three of the steps below are checked by the project's own build process rather than by memory. Running the quality gate is part of publishing, and it exits with an error — blocking the deploy — when a page is missing its research note, falls below the depth threshold, or lacks a lead and FAQ. Those are marked as enforced. The remaining steps are editorial judgement, and are described here plainly so you can hold the site to them.
Research before writing
The live search results for the target topic are pulled and read first — which pages currently rank, what format they take, how strong those domains are, whether an AI Overview sits on the query, and what people are asking alongside it. The findings go into a written research note kept in the project, one per page, with the decisions that follow from them: page type, title, and which questions the FAQ has to answer.
Enforced: the build fails if a published page has no completed research note.
Write the questions and the explanations together
Question banks are written by hand, one item at a time. An item is not finished until its explanation is written — why the correct answer is correct, and where a tempting wrong option comes from. The explanation is the part that makes a practice test worth taking, so it is never added afterwards.
Verify the answer keys mechanically
Every bank is run against its own key to confirm a perfect-answer pass scores 100%, which catches an off-by-one key or a mislabelled option. Correct answers are then checked for spread across option positions, so no test can be gamed by always picking the same slot. One early bank had every correct answer in the first position; that is exactly the error this step exists to catch.
Check depth and internal consistency
Counts claimed in a title have to match what the page actually contains, search volume has to justify the page existing at all, and each page needs a lead and a real FAQ rather than a stub. An early page was titled "60+" while holding around 39 items; it was expanded rather than renamed, and the check that would have caught it was added afterwards.
Enforced: the build fails on thin pages, missing FAQs, or a published page below the word threshold.
Cite the sources for anything claimed
Where a page refers to an established framework or an assessment instrument, it links to the primary source rather than naming it in passing — the test publisher, the standards body, or the original paper. Every link is checked to resolve before it goes live. Where the evidence is weaker than a claim would suggest, the claim is softened rather than propped up with a weak citation.
Publish, then keep it current
Each page shows a visible "updated" date and the name of the person responsible. Pages are revisited when the search results for their topic shift, when a bank is expanded, or when a correction comes in.
What we do not do
- No generated filler. Questions, explanations and guides are written for this site. Nothing is scraped, spun, or lifted from another test.
- No real exam items. Practice tests modelled on named exams such as the Wonderlic or the CCAT use original questions written to match the published format. Actual items from those exams are copyrighted by their publishers.
- No invented credentials. Names, qualifications, ratings and reviews on this site are real or absent. Where a figure or a norm does not exist, the page says so instead of estimating one.
- No claim of diagnostic validity. The tests have no norm sample and no published reliability or validity study, and the methodology page states that plainly rather than burying it.
Advertising and affiliate links
The site is ad-supported, and some pages carry affiliate links — currently to Amazon for books and games. Those links are marked where they appear, and the page carries a disclosure. A product is included because it fits the topic; the link is added afterwards, and never decides what gets recommended. Removing every affiliate link would not change which books are on the list.
Corrections
A wrong answer key in a practice test is worse than no practice test, so corrections take priority over new pages. If a question, key, explanation or fact looks wrong, report it and it will be checked against the source. Confirmed errors are fixed directly on the page, and the page's updated date changes with the fix.
Reach Christoph Ballmer at info@trainthinking.com.