Inductive reasoning tests: finding the rule

The skill

What is an inductive reasoning test?

An inductive reasoning test presents a small set of examples — a series of figures, a run of numbers, a group of items, two sets built on an unstated principle — and asks you to identify the regularity behind them and extend it. The rule is never written down. Recovering it from the evidence is the assessment.

SHL describes the task in its own example material as finding a logical sequence of five boxes and deciding which box completes the sequence, and its Verify G+ fact sheet describes the inductive questions as completely non-verbal, featuring only shapes and figures, measuring the ability to detect regularities, patterns and generalisations and infer rules that can be applied to different situations. Aon's scales ix shows nine objects where every object complies with a governing rule except one, and asks which does not comply.

Because the rule is inferred rather than given, an inductive answer is the best-supported hypothesis rather than a proof. The right discipline is to find the simplest rule that accounts for every example, not the most ingenious one that accounts for the step in front of you.

Try it before anything else

Free inductive reasoning practice questions

Six original questions in the shapes this guide teaches: figure series, odd-one-out groups, and the letter and number sequences that carry the same task in symbols. Commit to an answer before opening the solution — each one shows the scan in use and names why every wrong option is on the list.

What employers are looking at

What do inductive reasoning tests measure?

Employers use inductive items to predict how someone handles a problem with no supplied method: a new system, an unfamiliar dataset, a process nobody has documented. SHL states that individuals high in this ability tend to excel in global and strategic thinking and are good at finding errors in work processes.

Three abilities separate here. Hypothesis formation: can you produce a candidate rule quickly rather than staring at the figures. Testing: do you check the rule against every example before choosing, including the ones that already looked easy. Revision: when a rule fails on step three, can you drop it instead of bending it.

Reading your result

How inductive reasoning tests are scored

There is no single scale behind an inductive score. What the providers reviewed here publish is mostly about what the score is not — not a pass mark, and not a simple count of right answers — and knowing that changes how you should read your own result.

• No provider publishes a pass mark. None of the providers reviewed here publishes a pass mark for an inductive assessment, or a target number of questions. Korn Ferry states directly that it sets no pass or fail marks and that benchmarks belong to the hiring organisation. If a recruiter has disclosed a target, that is the only target worth working to.
• The number you answer correctly is not the score. Pearson states that because the questions in its DAT range carry different difficulty weightings, the number answered correctly is of little value when comparing scores. On an adaptive assessment two candidates with the same count may have faced very different questions, so the count alone settles nothing.
• On an open-ended test there is nothing to finish. Aon's scales clx creates a new question every time you complete one, and the test may get more difficult as you progress. The count of completed questions is not a figure you can reason about from the outside, because difficulty rises with your accuracy. Work at a rate you can sustain without guessing.
• Finishing is not the target on a timed set either. Aon tells candidates preparing for scales ix that the assessment is timed but they are not expected to answer all the questions, and that the aim is to get as many correct as possible in the time provided. Accuracy at a sustainable pace beats rushed completion.
• Difficulty is not feedback. SHL's Verify G+ is adaptive, Pearson's DAT follows a correct answer with a harder question, and Aon presents scales ix in increasing levels of difficulty. On tests built this way a harder question often follows a correct answer, so you cannot read your score from how hard the test feels.

Question formats

What do the questions look like?

Inductive assessments recur in a handful of shapes. Naming the shape decides which scan you run.

• Next in series. A row of figures changing step by step, with the next one to be chosen. SHL's published example format is a logical sequence of five boxes with the completing box selected from options A to E.
• Odd one out. A group of items where all but one obey a shared rule. Aon's scales ix presents nine objects and asks which one does not comply with the governing rule.
• Rule matching across sets. Two examples that follow a rule, and several candidates of which some follow the same rule. Aon's scales clx shows two three-by-three example grids and asks which two of four further grids obey the same rule.
• Alternating and interleaved changes. Two rules running at different rhythms, so a feature changes every step while another changes every second step. Reading only consecutive pairs hides this.
• Simultaneous independent attributes. Count, rotation, shading and position each changing on their own schedule inside one item. The scan matters more here than cleverness.
• Classification into groups. New figures assigned to one of two sets defined only by example. Aon files this shape under scales cls, which Cognivy teaches on the abstract reasoning guide.
• Number sequences. A run of numbers with one missing, where the operation between terms has to be worked out before anything can be calculated. Pearson's DAT Numerical Sequences module uses numerical content for the same rule-discovery task.

Launch coverage

Which providers use inductive reasoning tests?

Three of Cognivy's launch families publish an assessment taught here, and each defines the task differently. Match the exact product name in your invitation.

• Verify Interactive Inductive Reasoning. SHL's published example format is a logical sequence of five boxes with one to be completed from options A to E. SHL's Verify G+ fact sheet, which combines numerical, deductive and inductive questions, lists ten inductive questions that are completely non-verbal and feature only shapes and figures, inside a 36-minute adaptive test of 30 questions.
• scales ix. Aon's inductive reasoning module shows an image of nine objects where every object complies with a governing rule except one, and asks you to determine which. Aon lists 20 items presented in increasing levels of difficulty, at 5 minutes, and tells candidates they are not expected to answer all the questions.
• scales clx. Aon's inductive logical reasoning module presents two three-by-three example grids that follow a rule and asks which two of four further grids follow the same rule. Aon lists it at 6 minutes and states that there is no set number of questions, because the system creates a new one each time you finish.
• Abstract Reasoning. Saville's Abstract item is a horizontal figure series: one row of six to nine symbols with a gap to fill, answered from five picture options. It is taught here rather than on the abstract reasoning guide because a figure series is what it is, whatever the product is called; Saville publishes no matrix or odd-one-out format anywhere in its aptitude roster. Note also that Abstract sits in Swift Executive Aptitude, not Swift Analysis, whose third section is Diagrammatic with four options. Saville lists Swift Executive Aptitude at 18 minutes across numerical, verbal and abstract reasoning, and Swift Global Aptitude at 12 minutes across number reasoning, abstract reasoning and memory recall.
• DAT Abstract Reasoning. Pearson describes the subtest as looking at the ability to see sequences in shapes and patterns, delivered with computerised adaptive testing at 15 items in approximately 12 minutes. It is taught here rather than on the abstract reasoning guide because a figure sequence is what the task is: that guide's practice is odd-one-out, which asks a different question of the same material. The abstract reasoning guide still describes the product, because abstract material is what it is made of.
• Logiks Abstract (Advanced). Available as a standalone ability or as one section of Logiks General. Talogy publishes the range at 12 to 25 minutes and 20 to 50 items depending on version, drawn from an item bank so that no two assessments match. Taught here for the same reason as Pearson's DAT Abstract: Talogy's own summary of the task is completing figure sequences, and this is the guide backed by a figure-series bank.
• DAT Numerical Sequences. Pearson describes this DAT Next Generation module as measuring problem solving abilities with numbers, and describes the neighbouring Numerical Calculations module as measuring addition, subtraction, division, percentages, algebra and multiplication. Cognivy teaches sequences here because the operation between the terms has to be discovered before any arithmetic can start, which is the inductive task in numerical dress; that split is Cognivy's classification of the task rather than a Pearson label. Pearson lists each module at 15 items and around 12 minutes, delivered untimed and unsupervised using computerised adaptive testing.

Provider names identify the assessment format. Cognivy is independent and is not affiliated with or endorsed by these providers.

What varies, and what does not

Timing, adaptivity and scoring

Inductive assessments are usually short, and the pacing model differs sharply between an item-limited and an open-ended test. These are the published figures.

• Aon scales ix: 5 minutes, 20 items in rising difficulty. Aon states that the items are presented in increasing levels of difficulty, that there is only one correct answer per item, and that you can navigate between statements although working in the predefined order is recommended.
• Aon scales clx: 6 minutes, no fixed question count. Aon states that the system will continue to give you a new question every time you complete one, and that the test may get more difficult as you progress. There is nothing to finish, so pace is the whole strategy.
• SHL Verify G+: 36 minutes for 30 questions across three abilities. SHL's fact sheet allows 36 minutes for a maximum of 30 questions, ten for each of numerical, deductive and inductive, in a multiple-choice adaptive format designed for an unproctored environment.
• Pearson DAT Numerical Sequences: 15 items, about 12 minutes. Pearson lists those figures for each DAT Next Generation module, describes delivery as untimed and unsupervised, and states that a correct answer is followed by a harder item. It also states that because items carry different difficulty weightings, the number answered correctly is of little value when comparing scores.
• Finishing is not the target on Aon's modules. Aon tells candidates preparing for scales ix that the assessment is timed but they are not expected to answer all the questions, and that the aim is to get as many correct as possible in the time provided.

A repeatable approach

A scan that works under time pressure

Running the same scan every time, instead of looking for whatever stands out, keeps the search order independent of the figure in front of you. Five steps, in this order.

1. 1

   Scan attributes in a fixed order. Count, then position, then rotation, then shading, then size, then shape type. A fixed order stops you fixating on the largest object and missing the small one that carries the rule.

2. 2

   Describe the rule in words before looking at the options. Say it plainly: one dot is added and the arrow turns a quarter clockwise. A rule you cannot state in a sentence is usually a rule you have not found.

3. 3

   Test it against every example, including the easy ones. A rule that explains steps one to three and fails on four is not the rule. On an odd-one-out item, confirm that the rule holds for all the others before you accept the exception.

4. 4

   Prefer the simplest rule that fits. If a one-step change explains the sequence, do not build a two-step one. Item writers use simple rules combined, not complicated rules alone, and Aon's own material notes that many tasks are quite straightforward.

5. 5

   Eliminate on the rule you are most confident about. Apply your surest rule to all four options first. It usually removes two of them, and the remaining decision is a single comparison rather than a full solve.

What slows progress

Common mistakes, and the fix for each

• Tracking only the most obvious element. Fix: run the scan in the same order every time. The rule is frequently carried by the small feature in the corner, and a fixed order is the only reliable defence against a big shape stealing your attention.
• Choosing an option that fits the last step. Fix: verify your rule against the whole sequence before you compare options. A rule that works from tile 4 to tile 5 and nowhere else is a coincidence you have mistaken for a pattern.
• Inventing a complicated rule. Fix: test one-step changes first — add one, turn a quarter, alternate. Complexity in these items normally comes from several simple rules running at once, not from one intricate rule.
• Missing an alternating rhythm. Fix: when consecutive steps refuse to line up, compare every second figure instead. Alternating rules are invisible to a pairwise scan and obvious to a two-step one.
• Repeating familiar practice items. Fix: practise on unseen sequences. Reworking a set you have done before measures recall rather than rule discovery, so a score built on remembered answers tells you nothing about an item you have not met.

In practice

How do you find the rule when nothing jumps out?

Stop looking for whatever stands out and run the scan instead: count, then position, then rotation, then shading, then size, then shape type. The order is the point — a fixed sequence stops the largest shape stealing your attention, and the rule is frequently carried by the small feature in the corner. Rejecting an attribute explicitly, the way sample question 2 rejects count before it finds shading, is progress even when it feels like nothing happened.

Propose from one transition, then confirm on the rest. A single step — tile one to tile two — is enough to suggest a candidate rule, and stating it in a sentence is the test of whether you have actually found one. Then check it against every remaining step, including the easy-looking ones: a rule that explains three steps and fails the fourth is a coincidence, not a pattern.

When consecutive steps refuse to line up at all, change the comparison rather than the effort: read every second figure. Alternating rules are invisible to a pairwise scan and obvious to a two-step one — sample question 5 runs two alphabets through one row of letters, and no amount of staring at neighbouring terms will reconcile them.

In practice

Rotation or reflection — how do you tell?

Use the swing test, because it cannot lie. Stand at the corner where a figure's arms meet, point along the long arm, then swing round to point along the short arm, and note whether the swing runs clockwise or anticlockwise. Turning a figure within the plane never changes the direction of that swing; mirroring it always reverses it. One check, and it settles the question every time.

The cues candidates actually use are worse. An unfamiliar angle is not evidence of a reflection — a tile hanging upside down is still the same tile — and corner position proves nothing either: in sample question 4, two cards put the corner in the same place and only one of them follows the rule. The rule lives in the order of the arms, not in where the figure happens to sit.

Run the same check on the answer options. A mirrored option is the standard distractor under a rotating sequence, because it matches the expected answer feature for feature and differs only in the direction of the swing. Tile C in sample question 3 is built exactly that way, and it is the option that catches candidates who match the long arm and stop looking.

In practice

Should you read the answer options before you solve?

Yes. The options are evidence, and using them is legitimate rather than cheating. When three options share a feature and one does not, the differing feature is usually the rule the question is testing, which can hand you the right attribute to scan before you have consciously found it in the sequence.

Eliminate on the rule you are most confident about. Applying your surest rule to all the options usually removes two of them at once, and the remaining decision is a single comparison rather than a full solve. That matters most where the clock is tightest: Aon lists scales ix at 5 minutes for 20 questions, and states that the aim is to get as many correct as possible in the time provided.

The caution is the option that fits the last step only. An option can extend the final transition perfectly while breaking the rule everywhere else, so confirm your rule across the whole sequence before you let the options vote. Sample question 1's fourth option simply repeats the final tile, and it is what gets chosen when one rule is found and the second is never checked.

A clear route

How to prepare in the days you have

Aon's own guidance notes that improving inductive logical reasoning is not easy, so spend the short run-up on the parts of the task a fixed method can change.

1. 1

   First, learn the attribute list. Count, position, rotation, reflection, shading, size, shape type, overlap. Learn it as a list you can recite, because that is what turns into a scan under time pressure.

2. 2

   Next, work untimed and state every rule aloud. Force yourself to put the rule in a sentence before choosing. If the sentence is vague, the answer is a guess, and finding that out untimed is much cheaper than finding it out on the day.

3. 3

   Then practise the shapes separately. Series and odd-one-out reward different first moves. Series work forwards through changes; odd-one-out works by confirming a rule across the majority. Mixing them early hides which one is weak.

4. 4

   Add the provider's clock and interaction. A 5-minute, 20-item module and an open-ended module that keeps generating questions need different pacing. Rehearse against the model your invitation names.

5. 5

   On the last day, review by cause. Sort your misses into rule not found, rule found but not checked, and rule checked but option misread. Each has a different fix, and only the first one needs more practice items.

Cognivy uses your assessment date to choose the route rather than asking you to predict a study schedule. When you sit down to practise, you choose the session length that fits that day.

Pace

Getting faster without losing accuracy

Speed here comes from stopping early with confidence rather than from scanning faster. Four habits do most of it.

• Compare two adjacent figures, not all five. A single transition is enough to propose the candidate rule. Use the rest of the sequence to confirm it rather than to discover it, so the later figures are checked against one rule instead of read for a new one.
• Use the options as evidence. When three options share a feature and one does not, the differing feature is usually the rule the item is testing. Reading the options early is legitimate and fast.
• Set an abandon point and honour it. On a timed module where you are not expected to finish, one rescued item costs the two after it. Choose the best remaining option and move.
• Rehearse the interaction, not only the shapes. Selecting two grids rather than one, or navigating between items, costs seconds until it is familiar. Aon's scales clx asks for two selections per question and lets you unselect, which is a different rhythm from a single-choice item.

Keep exploring

Related test types and how they differ

Inductive reasoning overlaps with several test types by name. These are the practical differences.

Direct answers

Inductive reasoning test FAQs