Abstract Reasoning

What is abstract reasoning? Meaning and definition

A simple definition of abstract reasoning: thinking with general ideas rather than particular things. The American Psychological Association defines abstract thinking as "thinking characterized by the use of general ideas or concepts", and abstraction as "the formation of general ideas or concepts by extracting similarities from particular instances" (APA, abstract thinking; APA, abstraction). The word itself comes from the Latin abstrahere, which "meant 'to remove forcibly' or 'to drag away'" (Merriam-Webster): you drag the rule away from the examples it came in.

It usually works in two steps. First you notice what a set of examples has in common, which psychologists call inductive reasoning, "the form of reasoning in which inferences and general principles are drawn from specific observations and cases" (APA). Then you use the rule on something new. That second step is what makes it useful: abstract reasoning lets you deal with situations you have never met, because you can see what they share with ones you have.

Examples of abstract reasoning

• Number patterns: seeing that 3, 6, 12, 24 doubles each time, and predicting 48.
• Proverbs: understanding that "don't count your chickens before they hatch" is about any outcome that isn't certain yet, not about chickens.
• Analogies: seeing that a judge is to a court as a referee is to a match, because both apply rules to a contest.
• Algebra: using x to stand for any number, and working with the relationship rather than a specific value.
• Strategy: recognising that a new problem at work has the same structure as one you solved last year.
• Shape puzzles: working out that a shape rotates and gains a side at each step, and choosing what comes next.

Abstract vs concrete thinking

Neither style is "bad". Concrete thinking is how young children naturally think, and it is the right tool for many practical jobs. The APA notes that it is "characteristic of young children and may also be seen in people with schizophrenia and people who have suffered a brain injury, especially frontal lobe damage". That is why explaining proverbs is an old way of testing abstraction: the APA lists a proverb test "used most often to measure intelligence" (APA), and in a 2022 study, literal answers scored one point and abstract interpretations two (Kljajevic, 2022). The same principle appears in the Similarities questions of the verbal reasoning section of IQ tests.

Free abstract reasoning test: Set A or Set B?

This four-puzzle abstract reasoning test uses a classic format: two sets of boxes each follow a hidden rule. Work out both rules, then decide whether each test box belongs to Set A, Set B or neither. The trick is that the rule is usually one feature, and everything else is there to distract you.

How to crack these: check one feature at a time across all four boxes in a set: how many shapes, which kinds, how many are shaded, what is missing. Stop as soon as one feature is the same in every box of Set A and different in every box of Set B. Then check it against the test box.

How abstract reasoning is tested

Raven's Progressive Matrices

The best-known abstract reasoning test is Raven's Progressive Matrices, published in 1938 by the British psychologist John C. Raven. The APA describes it as "a nonverbal test of mental ability consisting of abstract designs, each of which is missing one part": "60 designs arranged in five groups of 12", with items in each group getting harder. It "is often viewed as the prototypical measure of general intelligence" (APA Dictionary), though researchers have also shown it is not a pure one (Gignac, 2015).

What does a matrix test actually measure? Patricia Carpenter, Marcel Just and Peter Shell analysed every item of the advanced version in 1990. They found nearly all the puzzles used just five kinds of rule, and that the number of rules in a puzzle on its own accounted for 57% of the variation in how often it was answered wrongly. The abilities that separated strong and weak solvers were "the ability to induce abstract relations and the ability to dynamically manage a large set of problem-solving goals in working memory" (Carpenter, Just and Shell, 1990).

The five rule types identified by Carpenter, Just and Shell (1990), drawn as simple examples.

You can try puzzles built on these rules in our pattern recognition test, which includes three matrix questions.

In IQ tests

Full IQ tests measure abstract reasoning mostly in their fluid reasoning sections. In the adult WAIS-5, Matrix Reasoning sits in the fluid reasoning domain, which Pearson describes as "verbal, visual, inductive, quantitative, and logical" (Pearson WAIS-5 overview). Verbal abstraction appears too, in the Similarities subtest. Our guide to fluid vs crystallized intelligence explains how these fit together, and our free IQ test has a full section of matrix-style fluid reasoning questions.

In admissions and job tests

Employers and universities also use abstract reasoning tests, sometimes under names such as inductive or diagrammatic reasoning. The UK's UCAT, taken by would-be doctors and dentists, had an Abstract Reasoning section of 55 questions in 13 minutes, about 14 seconds a question (UCAT 2019 technical report). It was withdrawn in 2025. The consortium's reasons are revealing: "Abstract Reasoning has lower predictive validity than the other sections", and "the test also has high coachability", with scores rising over the years as candidates practised (UCAT).

Abstract intelligence: a short history

The idea that intelligence is mostly about abstraction goes back to the start of intelligence testing. In 1920 the American psychologist Edward Thorndike proposed three intelligences: mechanical, social and abstract. "By abstract intelligence is meant the ability to understand and manage ideas and symbols, such as words, numbers, chemical or physical formulae, legal decisions, scientific laws and principles, and the like." He added that measurement had made the greatest progress in the case of abstract intelligence (Thorndike, 1920, Harper's).

A year later the psychologist Lewis Terman went further: "An individual is intelligent in proportion as he is able to carry on abstract thinking" (Terman, 1921). Early writing of this kind often came bundled with prejudiced claims about groups of people that have long been discredited, but the core idea lasted. The APA still lists abstract intelligence, "the intellectual ability to think in terms of abstract concepts", as a distinct ability (APA), and Raymond Cattell's later idea of fluid intelligence covers much the same ground.

The "seat" of abstract reasoning in the brain

Textbooks often call the frontal lobes the seat of intelligence and abstract thinking. Brain research supports a version of that, with an important correction: the frontal lobes do the job together with other regions.

• Integrating relations needs the prefrontal cortex. In a brain-imaging study using puzzles adapted from Raven's matrices, combining two relations at once "was specifically associated with bilateral rostrolateral PFC", the front and side of the prefrontal cortex (Christoff and others, 2001).
• Damage shows the same. Patients with prefrontal damage "exhibited a selective and catastrophic deficit in the integration of relations", while patients with temporal lobe damage, matched for overall IQ, did not (Waltz and others, 1999).
• But it is a network. A review of 37 brain-imaging studies proposed the parieto-frontal integration theory (P-FIT), in which reasoning depends on areas of the frontal and parietal lobes working together (Jung and Haier, 2007). In 80 patients with brain lesions, damage to frontal and parietal regions in this network "predicted fluid intelligence loss, whereas damage outside these regions was not predictive" (Woolgar and others, 2010).

How abstract reasoning develops and ages

Childhood. In Jean Piaget's theory, abstract thought belongs to the fourth and final stage of development, the formal operational stage, "from about age 11 to adulthood", when children "can also deal with abstract ideas and hypothetical situations" (OpenStax Psychology 2e). The APA dates it from "around age 12" (APA). Younger children can spot simple patterns, but reasoning about hypotheticals and pure symbols, the "abstract logic" of algebra, usually comes later.

Adulthood. Puzzle-style reasoning peaks early. In a study of adults tested on matrix reasoning and similar tasks, peak ages ranged from 22 to 27, and the gap between 18- and 60-year-olds was about 0.6 to 0.7 standard deviations for reasoning (Salthouse, 2009). Verbal abstraction ages differently: in a 2022 study of 333 people aged 18 to 89, older adults were better than young adults at explaining proverbs, and the skill was "well preserved in late adulthood, despite considerable age-related cortical atrophy" (Kljajevic, 2022). Experience, it seems, makes up for what speed takes away.

Can you improve abstract reasoning?

Scores on abstract reasoning tests are surprisingly changeable, which is both good and bad news.

• Whole populations have improved. Dutch military conscripts took the same 40-item Raven's test for decades. The share of 18-year-old men scoring above 24 out of 40 "rose from 31.2% in 1952 to 82.2% in 1981/1982", a gain of about 20 IQ points in one generation: sons against their own fathers (Flynn, 1987). This rise, the Flynn effect, was largest on reasoning tests like these. Our guide to average IQ covers it in more detail.
• Practice raises scores. Across 50 studies and more than 134,000 people, retaking a cognitive test raised scores by 0.26 standard deviations on average, more with coaching and when the same form was used (Hausknecht and others, 2007). That is why UCAT called its abstract section coachable.
• Education helps. Each extra year of education raised IQ by roughly 1 to 5 points, on every broad type of ability studied (Ritchie and Tucker-Drob, 2018).
• Brain-training games mostly don't. A major review found "extensive evidence that brain-training interventions improve performance on the trained tasks" but "little evidence that training enhances performance on distantly related tasks" (Simons and others, 2016).

The practical lesson: if you face an abstract reasoning test for a job or course, practise the format, because familiarity is worth real points. Just don't mistake a higher score after practice for a change in underlying ability.

Frequently asked questions

Sources

• American Psychological Association, APA Dictionary of Psychology: abstract thinking, abstraction, abstract intelligence, concrete thinking, inductive reasoning, proverb test, Raven's Progressive Matrices, formal operational stage.
• Merriam-Webster. Abstract.
• Thorndike, E. L. (1920). Intelligence and its uses. Harper's Monthly Magazine, 140, 227–235.
• Terman, L. M. (1921). Intelligence and its measurement: a symposium, II. Journal of Educational Psychology, 12(3), 127–133.
• Carpenter, P. A., Just, M. A. and Shell, P. (1990). What one intelligence test measures: a theoretical account of the processing in the Raven Progressive Matrices Test. Psychological Review, 97(3), 404–431.
• Gignac, G. E. (2015). Raven's is not a pure measure of general intelligence: implications for g factor theory and the brief measurement of g. Intelligence, 52, 71–79.
• Pearson. WAIS-5 overview.
• UCAT Consortium. UCAT 2025 changes; UCAT 2019 technical report, executive summary.
• Christoff, K. and others (2001). Rostrolateral prefrontal cortex involvement in relational integration during reasoning. NeuroImage, 14(5), 1136–1149.
• Waltz, J. A. and others (1999). A system for relational reasoning in human prefrontal cortex. Psychological Science, 10(2), 119–125.
• Jung, R. E. and Haier, R. J. (2007). The parieto-frontal integration theory (P-FIT) of intelligence: converging neuroimaging evidence. Behavioral and Brain Sciences, 30(2), 135–154.
• Woolgar, A. and others (2010). Fluid intelligence loss linked to restricted regions of damage within frontal and parietal cortex. PNAS, 107(33), 14899–14902.
• OpenStax. Psychology 2e, 9.2 Lifespan theories.
• Salthouse, T. A. (2009). When does age-related cognitive decline begin? Neurobiology of Aging, 30(4), 507–514.
• Kljajevic, V. (2022). Older and wiser: interpretation of proverbs in the face of age-related cortical atrophy. Frontiers in Aging Neuroscience, 14, 919470.
• Flynn, J. R. (1987). Massive IQ gains in 14 nations: what IQ tests really measure. Psychological Bulletin, 101(2), 171–191.
• Hausknecht, J. P. and others (2007). Retesting in selection: a meta-analysis of coaching and practice effects for tests of cognitive ability. Journal of Applied Psychology, 92(2), 373–385.
• Ritchie, S. J. and Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358–1369.
• Simons, D. J. and others (2016). Do "brain-training" programs work? Psychological Science in the Public Interest, 17(3), 103–186.