Free inductive reasoning test

What inductive reasoning is

The APA Dictionary defines inductive reasoning as drawing general principles from specific observations and cases (APA Dictionary: inductive reasoning). It is the reasoning behind forming a hypothesis from particular facts, which is why the APA also calls it a cornerstone of the scientific method. Deductive reasoning runs the other way: it shows that a conclusion follows necessarily from premises (APA Dictionary: deductive reasoning). Our page on inductive vs deductive reasoning sets the two side by side. This page is for testing and training the inductive half.

Induction sits close to the center of intelligence as researchers measure it. Intelligence research treats inductive reasoning tests as markers of fluid intelligence (Loe et al., 2018), and in a 2023 reanalysis of Woodcock-Johnson data, Kevin McGrew calls inductive reasoning the primary indicator of fluid reasoning (Gf) and of general intelligence (g). In that analysis the Concept Formation test, a clear measure of induction, had the strongest Gf loading, .54 (McGrew, 2023).

Every inductive item works the same way. You get a few examples, you guess a rule, and you test the guess against the examples before using it. Klauer's model of inductive reasoning describes six processes and is assessed with item types such as class formation, analogies, series completion and matrices (Pásztor et al., 2022). Analogies and series are the most typical formats (Primi et al., 2013). The formats look different, but the core move, compare cases and find what is constant and what changes, is identical.

Studies of how people solve matrix problems show what separates strong and weak solvers. Everyone used the same incremental strategy of encoding the figures and inducing the regularities one at a time; the differences lay mainly in the ability to induce abstract relations and to manage many problem-solving goals in working memory (Carpenter, Just & Shell, 1990).

How inductive reasoning is measured

Almost every major IQ battery measures induction under the heading of fluid reasoning. The WAIS-5 Fluid Reasoning index uses Matrix Reasoning and Figure Weights (Pearson: WAIS-5 overview). The Stanford-Binet 5 has a Fluid Reasoning factor (PRO-ED: SB5), and the Woodcock-Johnson uses Concept Formation (McGrew, 2023). The ETS kit of reference tests measures an Induction factor with a Letter Sets Test, a Locations Test and Figure Classification (ETS: Kit of Factor-Referenced Cognitive Tests).

Matrix tests are the best-known inductive measures. Carpenter and colleagues identified five rules in Raven's items: constant in a row, quantitative pairwise progression, figure addition or subtraction, distribution of three values, and distribution of two values. Constant in a row is the easiest to induce and distribution of two values the hardest (Lozano, Embretson & Revuelta, 2025). Raven's is a strong measure of general ability but not a pure one: across several large samples it shared about 50% of its variance with g (Gignac, 2015).

Matrix research also shows what makes an inductive item hard. Difficulty depends on both the type of rule, which taxes finding what corresponds to what, and the number of rules, which taxes goal management in working memory (Lozano, Embretson & Revuelta, 2025). People also learn while they take the test: items using the pairwise progression and figure addition or subtraction rules got easier as test takers went along, and how much a person learned this way correlated with outside measures of ability (Lozano, Embretson & Revuelta, 2025). Spotting a rule faster the second time is part of the skill being measured.

The Cattell Culture Fair test is induction from start to finish: its four subtests are series, classification, matrices and conditions (Zahra et al., 2022). School tests such as the CogAT measure reasoning in verbal, quantitative and nonverbal domains (Riverside: CogAT Form 8), and number series count too: in CHC research they are classed as quantitative reasoning, a narrow ability under fluid reasoning (McGrew, 2023). In hiring, the PI Cognitive Assessment reports an Abstract Reasoning subscore (Predictive Index: Cognitive Assessment). See types of IQ tests for how these fit together.

The ten questions above sample three inductive formats: four shape sequences, four 3x3 matrices and two odd-one-out sets, ordered from easy to hard. The odd-one-out sets are classification items: five figures share a rule and you find the one that does not.

Worked examples: easy, medium and hard

The easy example is a sequence with one rule. Compare the first two tiles, name the change, and confirm it on the next pair before you extend it. One confirmed rule is worth more than three guessed ones.

The medium example is a matrix that uses distribution of three values. Nothing progresses here. Instead, each shape and each fill appears exactly once in every row and column, so the missing tile is whatever combination is not yet used in its row and its column.

The hard example is a number series, the same inductive move with numbers. The numbers themselves look irregular, so look at the gaps between neighbors instead and ask how the gaps change. This format is not in the ten questions above, but it is the Quantitative Reasoning section of the full test.

Number series research describes the same steps you just used: detect the relations between neighbors, find the period of the pattern, describe it, and extrapolate (Loe et al., 2018). In one set of generated number series, five rule types together accounted for 69% of the variance in item difficulty (Loe et al., 2018), so the kind of rule, not luck, is what makes an item hard.

Common mistakes on inductive puzzles

• Accepting a rule that fits two examples. Any two tiles fit many rules. Test the rule on every example you have, including the full row and column of a matrix.
• Looking for a progression when there is a distribution. In distribution-of-three matrices nothing grows or turns. Each value simply appears once per row and column.
• Stopping at one rule. Harder items stack rules. If an option matches your rule but two options do, a second rule is deciding between them.
• Being distracted by surface features. In odd-one-out sets, size, tilt and fill vary on purpose. The rule is a side count, not how the figure looks.
• Reading the numbers, not the gaps. In number series the rule usually lives in the differences between neighbors. Write the gaps down.
• Looking at the options first. Predict the answer, then find it. Starting from the options invites you to pick whatever seems close.
• Running out the clock. Each question has a 30-second ring. If you have a rule that fits most examples at about 20 seconds, use it.

One routine works on every format. First, describe the examples without judging them: what shapes, fills, sizes, positions or numbers are there. Second, compare two examples and list every difference. Third, form the simplest rule that explains the differences and check it against every other example. Fourth, predict the answer before you look at the options. If no option matches, go back to step two; you missed a difference.

How to practice inductive reasoning

Inductive reasoning has one of the stronger training records of any reasoning skill. Klauer and Phye's meta-analysis covered 74 training experiments with nearly 3,600 children and confirmed gains in both fluid intelligence and classroom learning. The training effects on intelligence test scores grew over time, transfer to academic learning was larger than transfer to intelligence tests, and the authors reported that the results cannot be explained by placebo or test-coaching effects (Klauer & Phye, 2008). As reported by a later paper, the average corrected effect was d = .52 on intelligence tests and d = .69 on learning a new academic lesson (Pásztor et al., 2022). A plan that covers all the formats:

1. Sequences first. Work through the easy, medium and hard sequence puzzles. They teach you to compare neighbors and count rules.
2. Then matrices. The matrix library moves from row rules to distributions to line overlays. The matrix reasoning guide explains each rule.
3. Classification. The odd-one-out puzzles train you to state a shared rule before you name the exception.
4. Numbers. The number series library and the number series guide carry the same method into arithmetic.
5. Test each format on its own. The pattern recognition test, abstract reasoning test and number sequence test each drill one format, ten items at a time.
6. Review every miss. Read the worked solution and find the exact example your rule failed on. That habit is the inductive skill itself.

Two limits are worth knowing. Retaking a cognitive test raises scores by about a quarter of a standard deviation on average (Hausknecht et al., 2007), and gains level off after about the third sitting (Scharfen, Peters & Holling, 2018), so part of any quick rise is format familiarity. And reviews of cognitive training in general find minimal effects on domain-general skills once study quality is accounted for (Sala & Gobet, 2019). The inductive training studies are an exception worth noting, but they tested structured training programs with children, not ten minutes of puzzles.

Inductive reasoning on the full IQ Octopus test

The full test has 25 items in five sections: Pattern Recognition (5), Abstract Reasoning (9), Spatial Reasoning (5), Quantitative Reasoning (3) and Logical Reasoning (3). Most of them rely on induction: the sequences, the matrices, the number series and the figure analogies all hide a rule you have to find. Items are generated fresh for every attempt, so practice helps you know the formats but there is no answer key to learn.

A practice score out of ten is not an IQ. IQ is a deviation score that compares you with people your own age on a scale with a mean of 100 and a standard deviation of 15 (Wikipedia: Intelligence quotient). The full test scores all 25 answers with a Rasch model and reports an estimated IQ with a confidence range. Take the full IQ test when you are ready.