The question types in the order they tend to arrive — and how long each deserves.
Decision Making runs 35 questions in 37 minutes, and the question types tend to arrive in a rough order. Syllogism-style drawing conclusions usually come early — easy to medium, worth one to two minutes each. Logical puzzles follow — hard, and worth up to two and a half minutes when the diagram is paying off. Evaluating arguments are quick — medium-hard but only half a minute to a minute. Narrative-based conclusions are hard to very hard and can take up to three minutes; data-based conclusions are the most demanding and appear throughout.
Knowing this shape is itself a strategy: you can spend confidently on a puzzle because you know the evaluating-arguments questions ahead are cheap, and you can refuse to let a very hard data question take a fourth minute.
A strong argument is relevant, its premises actually support its conclusion, and it addresses both the problem and a workable solution. Weak options are usually true but irrelevant, restate the question, lean on an unstated assumption, or claim a causal link the premises do not earn.
Work by asking what would make this argument stronger or weaker, then test each option against that — rather than asking which option sounds most agreeable.
Narrative-based questions read like a compressed VR exercise: conclude only from explicit statements and clear implications, and never assume beyond the text. Data-based questions hand you tables, graphs or figures: establish what the data actually shows first, keep correlation firmly separated from causation, and watch for confounding factors the question is inviting you to ignore.
Exam-day DM tends to run harder and more abstract than practice banks. That is good news: the harder the test, the more it rewards preparation. The students it punishes are the ones who only ever drilled the friendly versions.
Real improvement here comes from working difficult official-style questions, understanding exactly why each wrong answer happened — an assumption made, wording misread, logic inverted — and then practicing that specific weakness until it stops recurring.