How our questions are made

Official exam questions belong to the exam bodies, so every question on 180 Prep is written for 180 Prep. That puts the burden on us to prove each one is fair. This is how we do it.

Built to the published format

Each exam runs in its exam body's own published structure: section lengths, timing, number of answer choices, and question types. An LSAT section and an MCAT section look nothing alike, because they aren't. Every section is checked against that specification before it ships, and ordered scales such as the Data Sufficiency rubric or the UCAT Situational Judgement ratings keep their official order.

One defensible answer

New questions are written by one AI model and then solved blind by a second model from a different company, which never sees the intended answer. If the two disagree, if the checker calls the question unfair, or if it rates the question easier than the real exam, the question is thrown away. In a measured run, about two in three drafts were rejected.

Where an answer can be computed from the question's own data — arithmetic, a table, a chart — a solver has to reproduce the credited answer or the question is rejected.

No shortcuts to the answer

A question bank can be free of typos and still be broken as a test: if the right answer is usually the longest choice, a student can score above chance without reading. We audit every bank for that and for answer position, and every bank sits at chance on both. The check runs on every release, and a release that drifts is refused.

When something slips through

Any question can be reported from inside the app. Reports are reviewed, and a question that fails review is fixed or removed.

What we don't claim

180 Prep is not a substitute for official practice tests — those are the benchmark, and we recommend them for full-length rehearsal. We don't guarantee scores. None of the exam bodies endorse, sponsor or are affiliated with 180 Prep.

Home · How our questions are made · Security · Privacy · Terms