Past papers
GMAT — Reading previous papers properly
Because the test is adaptive and administered year-round, there is no annual paper to work through. GMAC instead publishes official guides and practice questions drawn from retired test material, and these are considerably more valuable than third-party imitations — the logical precision of GMAT critical reasoning and data sufficiency questions is difficult to replicate, and material that gets it slightly wrong teaches slightly wrong habits.
How to use official material
- Work official questions before third-party ones. They define the standard of logical precision the test uses.
- For every critical reasoning question, articulate the argument’s structure. Premise, conclusion, assumption. Doing this explicitly is what builds the skill.
- For data sufficiency, resist solving. The question asks whether you could solve, not what the answer is. Solving wastes time and causes errors.
- Keep an error log by question type and cause. Adaptive scoring means recurring error types are expensive.
- Take official practice tests under real conditions. Adaptive practice tests are the only realistic score predictor.
Why official questions are worth more than third-party ones
- The distractors are constructed to a standard — GMAT wrong answers are designed to be attractive for specific reasons — a scope shift, a reversed causal claim, an unstated assumption. Third-party questions frequently miss that construction, which means practising on them teaches you to spot the wrong things.
- The logical precision is consistent — official critical reasoning questions have exactly one defensible answer. Questions that do not meet that standard train tolerance for ambiguity, which is the opposite of what the test rewards.
- The difficulty calibration is real — official material is labelled by difficulty against actual candidate performance rather than by an author’s judgement.
- The register of the passages is right — reading comprehension passages have a characteristic density and subject range that third-party material approximates unevenly.
For every critical reasoning question, articulate the argument’s structure before looking at the options: what is the conclusion, what supports it, and what gap sits between the two. Candidates who go straight to the options are choosing between five plausible sentences; candidates who have named the gap are checking which option addresses it. The second process is slower for the first few weeks and considerably faster and more accurate thereafter.
For data sufficiency, the discipline is to resist solving. The question asks whether the statements would suffice, not what the answer is — and computing the answer is both unnecessary and a reliable way to run out of time. Practise stopping at the point where sufficiency is established, which feels wrong at first and is the entire technique.
An error log by type and cause, not by topic
At this level, the useful classification is what went wrong rather than what the question was about: misread the scope, accepted an unstated assumption, computed unnecessarily, or lost the thread of a long passage. Those are the failures the GMAT is built to expose, and naming them is how they get corrected.