Nearly every intelligence assessment ever built draws on the same five families of question. Learn what each family is measuring and the traps built into it, and an unfamiliar paper stops being a surprise. That is the whole benefit of preparation: not knowing answers in advance, but never losing time working out what is being asked.
This page walks through each format with worked reasoning, sets out the errors that cost the most points, and explains why memorising answer keys is the one preparation strategy that reliably makes a result worse.
Matrix Reasoning
A grid of shapes with one cell missing, and several candidate answers below it. This format dominates modern assessment because it needs no language, travels across cultures better than any alternative, and loads heavily on fluid reasoning - the ability to work out a rule you have never been taught.
Rules operate along rows, down columns, and often both at once. The common ones are worth knowing by name: progression, where a feature increases step by step; rotation, where an element turns by a fixed angle; addition and subtraction, where two cells combine to produce the third; and distribution, where every element appears exactly once in each row and column.
The reliable method is to fix one variable at a time. Ignore shape and track only colour across a row. Then ignore colour and track only count. Then position, then orientation. Isolating attributes turns a bewildering grid into three or four simple observations that combine into the answer.
The trap is the distractor set. Wrong options are constructed deliberately: one satisfies the row rule but breaks the column rule, one is correct but mirrored, one is right in every respect except a single detail. An answer that looks obviously correct at a glance is frequently the near-miss. Confirm every rule you identified before committing.
Number and Letter Series
A sequence with a gap, usually at the end. The task is inducing the rule that generates it.
Begin with first differences - subtract each term from the next. If those differences are constant, the rule is arithmetic. If they form a recognisable pattern themselves, the rule is quadratic or built from a second layer. If differences lead nowhere, try ratios for a geometric rule, then check whether alternate terms form two interleaved sequences, which is a favourite construction at the harder end.
Letter series work identically once letters are converted to positions in the alphabet. Writing the numbers above the letters takes ten seconds and makes the structure visible immediately. Some series use alphabet position, some use distance between letters, and a few use both across interleaved strands.
The commonest error is overfitting. Given 2, 4, 8 the intended rule is almost always doubling rather than something involving cubes and remainders that also happens to fit. Assessments are built for the simplest rule that accounts for every term. If your explanation requires three clauses, it is probably wrong.
Verbal Analogies and Vocabulary
Doctor is to hospital as teacher is to what? These load on crystallised intelligence - accumulated knowledge - rather than fluid reasoning, which is why they advantage wide readers and handicap anyone working in a second language.
The method is to state the relationship in a full sentence before looking at the options: "a doctor works in a hospital". Then apply the identical sentence to the second pair. Vague relationships produce vague matches, and vague matches are where the distractors live.
Relationship types recur: part to whole, tool to user, cause to effect, category to member, degree of intensity, and opposition. Naming the type narrows the field fast. Where two options both seem to fit, the relationship was stated too loosely - tighten the sentence and one of them will fail.
Vocabulary items are more straightforward but reward structural knowledge. Prefixes, suffixes and Latin or Greek roots let you place an unfamiliar word approximately, which is usually enough to eliminate two options and turn a guess into a reasonable bet.
Spatial Rotation and Folding
Two objects shown from different angles: are they the same object rotated, or mirror images? Or: here is a flat net, which of these cubes could it fold into?
Rotation items reward a fixed procedure. Choose one distinctive feature on the first object - an odd corner, a marked face - locate it on the second, then check whether a single feature adjacent to it sits on the same side. Mirror pairs betray themselves at that second check, because handedness cannot survive rotation.
Folding items are best approached by elimination rather than visualisation. Take one face of the net and work out which faces end up adjacent to it and which end up opposite. Opposite faces on a cube can never both be visible, so any option showing two opposite faces together is out. Two or three such checks usually leave one candidate standing.
Response time is the trap here. Spatial items feel solvable by staring, and staring consumes the clock. A systematic check beats intuition on anything past the easy items, and the difference compounds across a speeded paper - one reason a well-constructed iq test reports timing alongside accuracy.
Odd One Out and Classification
Four or five items, one of which does not belong. Deceptively simple, and the source of more disputed answers than any other format.
The difficulty is that several defensible groupings usually exist. Given circle, square, triangle, cube, the answer is the cube because everything else is two-dimensional - but a case can be made for the circle as the only shape without straight edges. Well-built items eliminate the ambiguity by making one property shared by all but one and every other property mixed. If you find two valid answers, you have missed a property that separates them.
Work through attributes systematically: dimension, count, symmetry, colour, category, function. The intended rule is the one that leaves exactly one item outside the group with no competing rule doing the same job.
Format, Skill and Trap at a Glance
| Question family | What it measures | Best approach | Costliest trap |
|---|---|---|---|
| Matrix reasoning | Fluid reasoning, rule induction | Isolate one attribute at a time | Near-miss distractors |
| Number and letter series | Pattern induction, numeracy | First differences, then ratios | Overfitting a complex rule |
| Verbal analogies | Crystallised knowledge, relations | State the relationship as a sentence | Defining the relation too loosely |
| Spatial rotation | Mental transformation | Anchor feature, then check handedness | Mirror images read as rotations |
| Odd one out | Categorisation, abstraction | Run through attributes in order | Stopping at the first plausible rule |
Timing and How Many Items
Most standard batteries run 30 to 60 items across 20 to 40 minutes, which works out at roughly 40 seconds per item on average. That average is misleading. Early items should take ten seconds and late ones two minutes, so pacing yourself evenly guarantees running out of time exactly where the discriminating items are.
The practical rule is to bank time early. Move briskly through anything that resolves quickly, and treat the saved minutes as budget for the difficult end. Where the format allows revisiting, mark and skip rather than persisting - a stubborn item can consume the time three later ones needed. And find out before starting whether wrong answers are penalised, because if they are not, an educated guess on an unfinished item is free.
Practice Versus Memorising
Familiarity with formats genuinely improves scores, mostly by removing the time spent decoding instructions and the anxiety of meeting something new under a clock. That gain is legitimate and it is available to anyone with an hour to spend.
Memorising answer keys is a different thing and it is self-defeating. It inflates a number without touching the ability the number is supposed to represent, and it leaves you with a figure that will not survive any supervised setting - which is exactly why admission routes such as the Mensa IQ test use proctored papers and a single attempt. Drilling one specific item set also teaches that set rather than the underlying skill, so the gain evaporates the moment the format changes.
The sensible middle is variety. Rotate between matrix sets, series, verbal and spatial work so you are training reasoning rather than rehearsing a product. Sit occasional full-length papers under proper conditions, which is covered in the material on taking a free IQ test online without corrupting the result, and check that whatever you are practising on is worth the time by the standards set out for an accurate IQ test. For children, format familiarity matters even more than it does for adults, and the reasons are set out in the material on the IQ test for kids.
Understanding why an answer is correct is where the durable benefit lives. That habit - working out the rule rather than recognising the picture - transfers to every situation where several constraints have to be held at once, which is a fair description of most decisions worth thinking about. It is the same reasoning that runs through the material published by Virgin Games on probability, pattern and judgement under time pressure.
More on IQ Testing
Continue with the rest of the material on assessment, scoring and question formats:
