Decision Making is the most learnable section on the test: every question type has one clean method, and once you drill that method the question becomes almost mechanical. You get roughly a minute per question — comfortable compared with everything else — so the marks are lost to careless reasoning, not the clock. Treat accuracy, not speed, as the thing you are optimising, and decide your method before you read the options.
1 · Syllogisms — draw the cautious Venn, default to No
Sketch the smallest, least generous diagram the statements actually force, then test each conclusion against it. A conclusion only follows if it holds in every diagram you could legally draw, so when a statement could go either way, your default answer is No. Most wrong answers come from assuming an overlap the text never guaranteed.
2 · Strongest argument — must hit problem and solution
The winning option is relevant: it engages with both the problem in the stem and the specific solution being proposed, and explains how or why that solution works. Reject anything that is true but off-topic, that only restates the question, or that argues a different point entirely. Use the minor tie-breakers — specificity over vagueness, evidence over opinion, sensible and defensible in tone — only to separate two options that are both genuinely relevant.
3 · Venn and set counting — fill from the inside out
Place the figure the data pins down most tightly first, usually the central overlap, then work outward so every person or item is counted exactly once. Subtract known regions from totals rather than guessing, and read carefully for whether a number means 'only this set' or 'this set including overlaps'.
4 · Probability — fix AND vs OR, then convert
Decide whether outcomes combine with AND (multiply) or OR (add), and put every value into the same form — all fractions, or all decimals, or all percentages — before you touch them. For 'at least one', it is almost always faster to work out 1 minus the probability of none. A quick check that your probabilities sit between 0 and 1 catches most slips.
5 · Work the attack order, leave puzzles last
Spend your generous time where it converts best: clear syllogisms and strongest-argument items first, then Venn and probability, then the narrative and data 'drawing conclusions' sets. Save logical puzzles for the end — they are the biggest time sinks, so a grid you build slowly with full marks beats two rushed errors elsewhere.