Abstract reasoning tests
Abstract reasoning is what you use on a problem you have never seen before, when there is nothing stored to look up and the only way through is the rule you can work out on the spot. The free online abstract reasoning tests below put that to work in your browser, explain every answer as you go, and this page says why the result is not an IQ.
What abstract reasoning is
Abstract reasoning is problem-solving on material chosen to carry no information: a row of numbers, a row of letters, two pairs of ordinary words. There is nothing in an item to look up and nothing to have revised, so the only route to the answer is a rule you work out while you are looking at it. That is what puts the same handful of formats on employers’ aptitude tests, on school entrance papers and inside reasoning batteries — three audiences with almost nothing else in common.
Psychology calls the same thing fluid reasoning, and the distinction behind that name is Raymond Cattell’s. Fluid ability is reasoning on novel material; crystallized ability is the knowledge and skill you have accumulated. A vocabulary question measures what you were taught, a number series measures what you can work out. Most real thinking uses both together — but only one of them is what these pages are about.
In practice a fluid task is built so that nothing you already know can help. The terms are ordinary numbers or letters, the rule behind them is simple once you see it, and the whole difficulty is in seeing it. That is also why this family of task travels across languages and schooling better than most: there is nothing in it to have been taught.
What it is not: an IQ score. Series and matrices are the formats IQ tests are built from, and that is precisely why a browser version of one is not an IQ test — an IQ comes from a standardized battery, administered under controlled conditions and scored against a normative sample. None of those three things happens here, and NEURA publishes no norms to score anything against.
What these tests are not
- Not an IQ score
- The format is borrowed from IQ testing; the result is not an IQ. Nothing here is standardized, timed under controlled conditions or referenced to a sample, and NEURA never labels a number an IQ anywhere — on this site or in the app.
- Not a measure of what you know
- Nothing in these tests rewards education or vocabulary, by design. That makes them a poor way to measure anything you learned — and it is the reason somebody who has never met a number series can still solve one on the second attempt.
- Not a percentile
- No population comparison appears here, because we hold no representative sample to make one from. A run tells you how many rules you found out of the ones you were shown, which is a fact about the run and about nothing else.
How abstract reasoning is measured
The idea that a problem can be posed with no knowledge inside it is about a century old, and the two formats that made it work — a series of terms, or a grid of figures — are still the ones in use.
- 1904
One factor, many tests
Charles Spearman notices that people who do well on one mental test tend to do well on the others, and proposes a single common factor behind them. Everything after this is an argument about what that factor is, and about how to reach it without a school curriculum in the way.
- 1938
Two ways to pose a new problem
L. L. Thurstone argues that intelligence is several primary abilities rather than one, and uses letter series among the tasks for his reasoning factor — five or six letters, one rule, name the next. The same year John Raven publishes the Progressive Matrices, which asks the same kind of question with figures instead of symbols.
- 1943
Cattell names the split
Raymond Cattell proposes that general ability comes in two kinds: fluid, which handles problems you have not met, and crystallized, which is what experience has deposited. The pair became the backbone of nearly every modern model of cognitive ability.
- 1963
A program solves them
Simon and Kotovsky write a computer model that induces the rules behind Thurstone-style letter series. It made the task a favorite of cognitive science: to program it you must first say exactly what a person does, step by step.
- Today
Generated, not written
Series like these come out of an algorithm from a handful of parameters, which is how a test can hand out a fresh set every time without anyone authoring a single item. The sets on this site are drawn in your browser by ports of the generators NEURA uses.
Free abstract reasoning tests you can take here
Four abstract reasoning tests are published so far — Number Series Test, Letter Series Test, Number Analogy Test and Verbal Analogy Test. Each is free, runs online in this browser, and shows your own result the moment you finish.
None of them has a time limit, and every one runs client-side: no signup, nothing uploaded, and a plain-language reading of your result the moment you finish. The label on each card — fluid intelligence — is what psychology calls this ability; abstract reasoning is the same thing under the name the tests are sold under.
How to read your result
These tests report how many rules you found, out of how many you were shown — and how long you took over the ones you got. Four things worth knowing about that number.
-
Right answers, not speed
Neither test imposes a time limit, on purpose. A series has one defensible answer, and the interesting question is whether you found it — the clock is there so the result can tell you how long you took, not to hurry you.
-
The comparison that works
Your run against your own earlier runs. Every set is generated fresh, so two runs are never the same items — which is what keeps them comparable to each other instead of turning into a test of what you remembered.
-
A short set is a taste
A serious reasoning measure runs far longer and mixes formats, because any single format rewards whichever trick you happen to have. Treat a run here as a taste of the format, not as a measurement of your reasoning.
-
It is not a diagnosis
Nothing here is a clinical or educational assessment, and no decision about anybody should rest on it. It is a puzzle set with an honest score attached.
Abstract reasoning in the NEURA app
NEURA on iPhone works this domain on purpose rather than by accident. One of the 19 games is built for it, four more touch it while measuring something else, and the daily session carries four task formats that land there.
- Games built for it
- Tower Moves
- Games that also work it
- Equation Rush
- Math Sprint
- Pattern Matrix
- Cube Twist
- Daily-session tasks
- Number series
- Numeric analogies
- Verbal analogies
- Number matrix
A daily session is 20 tasks, about five minutes, and every answer is scored into the domain its task belongs to. Those domain results roll up into one number, which opens once you have completed three daily sessions — see Focus Index for what that number is and what it is not. A game’s primary domain takes the full weight of its result and the domains it also touches take a share. The exact weighting is not published, because the useful reading is the direction your number moves over weeks — not the arithmetic that made it.
Questions about abstract reasoning
What is an abstract reasoning test?
An abstract reasoning test poses problems you cannot have studied for. The material is ordinary — numbers, letters, everyday words, simple figures — and the task is to find the rule that generated it and apply that rule once more. Series, analogies and matrices are the classic formats, and all of them are scored on how many rules you found rather than on anything you knew in advance.
Abstract reasoning, fluid intelligence, nonverbal reasoning — are they the same thing?
Near enough, and the differences are worth knowing. Fluid intelligence is the term from the research literature, Raymond Cattell’s, and it names the ability rather than any one test. Abstract reasoning is what test publishers and employers call it, and the name this page is written under. Nonverbal reasoning is narrower: it means items with no words in them at all — figures, shapes and matrices, the format school entrance tests use. Of the tests on this page the two series sets and the number analogies are wordless in that sense; the verbal analogies are not, and the figure format sits on our Odd One Out Test, which measures pattern logic and has its own page.
Is an abstract reasoning test an IQ test?
No. The formats overlap — series and matrices are what IQ batteries are built from — but an IQ score requires a standardized battery, a controlled administration and a normative sample to score against, and a web page has none of those. NEURA publishes no norms and never labels any result an IQ, here or in the app.
Which test should I take first?
Take the Number Series Test if you think in quantities: the rules are arithmetic, and the jump from one term to the next is usually something you can say out loud. The Letter Series Test is the same idea in positions rather than amounts — the rules move through the alphabet, and knowing your way around it counts for more than arithmetic does. The two analogy tests ask a different question: instead of continuing one sequence, you work out what links a pair and apply that link to a second pair — the Number Analogy Test with quantities, the Verbal Analogy Test with everyday words. Series first if you want the shortest route in; analogies if you want the harder half of what reasoning batteries actually contain.
Can abstract reasoning be improved?
Practice makes you better at the abstract reasoning format, quickly and reliably — you learn where to look and which rules are common. Whether that improvement reaches beyond the format into general reasoning is the most argued-over question in the field, and the honest summary is that the evidence for broad transfer is weak. NEURA is built as a daily habit with a measurement attached, not as a promise about your reasoning.
Do I need an account to take these tests?
No. Every abstract reasoning test here is generated and scored in your browser: no signup, no upload, and no record kept once the tab closes. The app is the only place results are stored, and that is only so a trend across weeks can exist at all.
Finding the rule once is a moment. Finding one every day is training.
A browser run measures one sitting. NEURA on iPhone builds a fresh reasoning session every day, adapts the difficulty after every answer, and turns the runs into a Focus Index you can watch move over weeks. Your first session is free.
