IQ test questions fall into a handful of families, each aimed at a different broad ability: matrix and pattern problems for fluid reasoning, analogies and vocabulary for verbal reasoning, number series and word problems for quantitative reasoning, rotation and assembly tasks for visual-spatial ability, span tasks for working memory, and rapid scanning tasks for processing speed. The examples below were written for this article, not taken from any published test, so you can see how each type works without spoiling a real one.
| Domain | Typical question types | What it draws on |
|---|---|---|
| Fluid reasoning | Matrices, series completion, odd one out | Solving unfamiliar problems |
| Verbal reasoning | Analogies, vocabulary, similarities, comprehension | Language and acquired knowledge |
| Quantitative reasoning | Number series, arithmetic word problems | Reasoning with numbers |
| Visual-spatial reasoning | Mental rotation, paper folding, block assembly | Manipulating images in your head |
| Working memory | Digit span, letter–number sequencing, complex span | Holding and transforming information |
| Processing speed | Symbol search, coding, visual matching | Fast, accurate simple decisions |
These domains come from the Cattell–Horn–Carroll (CHC) framework, the most widely used map of human cognitive abilities. They overlap: nearly every item draws partly on general ability as well as on its own domain, which is why scores across very different question types correlate. That shared component is described in the g factor explained.
Matrix and pattern questions (fluid reasoning)
Matrix problems show a grid of figures with one cell missing. You infer the rules that govern the rows and columns, then pick the option that completes the grid. They are popular because they need little language or schooling. A classic analysis of matrix tests by Carpenter, Just and Shell (1990) found that the hardest items are hard mainly because they require discovering and holding several rules in mind at once.
Example. Each cell of a 3 × 3 grid contains a number of identical shapes:
- Row 1: 1 circle | 2 squares | 3 triangles
- Row 2: 2 triangles | 3 circles | 1 square
- Row 3: 3 squares | 1 triangle | ?
Which belongs in the missing cell: (a) 2 circles, (b) 3 circles, (c) 2 triangles, (d) 1 circle?
Answer: (a) 2 circles. Every row and every column contains each shape once and each count (1, 2, 3) once, so the last cell must be the missing shape, a circle, in the missing count, 2.
Analogies and other verbal questions (verbal reasoning)
Verbal items test how well you understand words and the relationships between ideas. Common formats include analogies, vocabulary definitions, "how are these two things alike" similarities and general comprehension questions. Because they depend on language and learning, verbal scores reflect education and reading more than matrix scores do. The CHC framework calls this kind of ability crystallized intelligence.
Example. Thermometer is to temperature as odometer is to: (a) speed, (b) distance, (c) fuel, (d) time.
Answer: (b) distance. A thermometer measures temperature and an odometer measures distance traveled; speed is what a speedometer measures.
Number series and word problems (quantitative reasoning)
Quantitative items test reasoning with numbers rather than calculation speed. Number series ask you to find the rule that generates a sequence; word problems ask you to set up a calculation from a short scenario. Most need nothing beyond basic arithmetic. The difficulty lies in spotting the structure.
Example. What number comes next: 3, 4, 7, 11, 18, 29, ?
Answer: 47. Each number is the sum of the two before it, so the next is 18 + 29 = 47.
Mental rotation and spatial questions (visual-spatial reasoning)
Spatial items ask you to manipulate a shape in your head: turn it, fold it or fit pieces together. The classic laboratory version comes from Shepard and Metzler (1971), who found that the time people need to decide whether two drawings show the same object rises steadily with the angle between them, as if they were rotating a mental image. Test items usually add a twist: some options are mirror images, which no amount of rotation can match.
Example. Picture a lowercase letter p in a standard typeface. Turning it within the page, without flipping it over, which of these can it become: (a) b, (b) d, (c) q?
Answer: (b) d. A half-turn (180°) of p gives d; b and q are mirror images of d and p, so reaching them would require a flip.
Span tasks (working memory)
Working-memory items ask you to hold information briefly and do something with it. Digit span is the simplest form: you hear a list of numbers and repeat it, sometimes backward or in sorted order. Research versions called complex span tasks interleave remembering with a second task, such as checking simple equations between the items to be recalled (Conway and colleagues, 2005). Working-memory capacity is substantially related to general reasoning, though a meta-analysis by Ackerman, Beier and Boyle (2005) concluded that the two are not the same construct.
Example. Read this sequence once, look away, then say it backward: 7 2 9 4 1 6.
Answer: 6 1 4 9 2 7. Reversing the list forces you to hold all six digits and work through them, which is harder than repeating them in order.
On a page this is only a sketch. In a real test the sequence is spoken or shown once and then removed, and the lists grow longer until you start making errors.
Speed tasks (processing speed)
Processing-speed items are simple on purpose. Almost nobody would get them wrong given unlimited time; in clinical tests the score is how many you complete correctly within a fixed limit, often a couple of minutes. They measure how quickly and accurately you can scan, compare and decide.
Example. How many times does the digit 7 appear in this row? 3 7 1 7 7 2 9 7 4 1 7 8
Answer: 5. On a real speed test you would face dozens of rows like this against the clock, and the challenge is keeping pace without errors.
Why IQ tests are timed
Most clinical tests time at least some subtests, for three reasons:
- Speed is one of the abilities being measured. For processing-speed tasks, the time limit is the test. Without it, nearly everyone would score near perfect.
- Standardization. Norms are valid only if everyone works under the same conditions. If the norming sample had 30 seconds per item and you had unlimited time, your score would be compared against the wrong baseline.
- Practicality. Time limits, together with rules that stop a subtest after a run of failed items, keep a session to a manageable length.
Timing has costs too. Psychometricians distinguish speed tests, where items are easy and the limit is tight, from power tests, where items get harder and time is generous. Tight limits on reasoning items can penalize careful, accurate people and those with anxiety, slower motor responses or a disability. That is why extended-time accommodations exist in many testing settings, though a score earned with extra time cannot be compared directly with timed norms.
Our test is untimed. Its processing-speed items are scored for accuracy only, so on this site that domain reflects careful visual checking more than raw speed, which is a real difference from clinical speed subtests. Answers given implausibly fast are flagged as low-confidence. You can take the free test at your own pace.
Does practicing sample questions help?
Seeing a few examples so the formats are not a surprise is reasonable, and real tests do the same by starting each subtest with practice items. Drilling hundreds of questions is different. A meta-analysis by Hausknecht and colleagues (2007) found that simply retaking a cognitive ability test raised scores by roughly a quarter of a standard deviation, about 4 IQ points, over intervals too short for any real change in ability. Heavy practice makes your result less accurate, not more. How scores can and cannot genuinely change is covered in can IQ change over time, and sensible preparation, mostly sleep and focus, is in how to prepare for an IQ test.