Watch your confidence skyrocket

Psychometric tests are standardised assessments that measure how you think, reason and behave. Employers use them because CVs and interviews only tell part of the story: a well-designed test gives an objective, like-for-like way to compare candidates on the skills a role actually demands, from interpreting data to making sound judgement calls under pressure. If you are applying for a graduate scheme, an apprenticeship, a professional role or a promotion, there is a good chance a psychometric test sits somewhere in the process.

The good news is that performance on these tests improves with practice. Familiarity with the question formats, the time pressure and the common traps means you spend your energy on the questions themselves, not on working out how the test works.

This page is your starting point. Below you will find every major test type explained, with a worked example question for each, guidance on how to prepare, and a plain-English explanation of how your score is actually judged. You can try an example of any of our tests for free, with no sign-up required, so you can see exactly what you are dealing with before you commit to anything.

The main types of psychometric test, explained

Most psychometric assessments fall into two families. Aptitude tests measure ability: they have right and wrong answers, they are usually timed, and they are scored against other candidates. Personality questionnaires measure preference and style: there are no right answers, and they are used to judge fit rather than skill. The sections below cover the types you are most likely to face, what each one measures, and what a typical question looks like.

Numerical reasoning tests

Numerical reasoning tests measure how confidently you work with numbers in a business context. They are not advanced maths exams. The underlying operations are things you learned at school: percentages, ratios, fractions, averages and currency conversions. The challenge is applying them quickly and accurately to data presented in tables, graphs and charts, under strict time pressure.

Expect somewhere between 15 and 30 questions, typically with around a minute to 90 seconds per question. Each question usually gives you a chart or table, asks you to extract the right figures, and then perform one or two calculations. Multiple-choice answer options are designed to catch common errors, so working carefully matters as much as working fast.

Example question. A company's revenue was £1.2 million in the first quarter and £1.5 million in the second quarter. What was the percentage increase in revenue?

Answer: 25%. The increase is £1.5m minus £1.2m, which is £0.3m. Divide the increase by the original figure: 0.3 divided by 1.2 equals 0.25, or 25%. The most common mistake is dividing by the new figure (1.5) instead of the original, which gives 20%, and you can be sure that 20% will appear among the answer options.

Try a free numerical reasoning test, no sign-up needed: Numerical Reasoning

Verbal reasoning tests

Verbal reasoning tests measure your ability to read a dense passage of text, understand precisely what it does and does not say, and draw accurate conclusions from it. Employers use them for any role where you will need to digest reports, contracts, emails or policy documents and act on them correctly.

The most common format gives you a short passage followed by a series of statements. For each statement you choose True (the passage supports it), False (the passage contradicts it) or Cannot Say (the passage does not give you enough information either way). The discipline is to judge each statement only on what is in the passage, setting aside anything you happen to know or assume about the topic.

Example question. Passage: "The company introduced flexible working in 2023. Since then, staff turnover has fallen and the majority of employees now split their week between home and the office." Statement: "Flexible working caused the fall in staff turnover." True, False or Cannot Say?

Answer: Cannot Say. The passage tells us both things happened, and in what order, but it never states that one caused the other. The turnover fall could have other explanations. Confusing "happened after" with "happened because of" is the single most common error in verbal reasoning tests.

Try a free verbal reasoning test, no sign-up needed: Verbal Reasoning

Logical and diagrammatic reasoning tests

Logical reasoning tests measure your ability to spot rules, patterns and relationships and apply them to reach a sound conclusion. Diagrammatic reasoning is a close cousin: instead of statements or sequences, you work with flowcharts and diagrams where symbols represent operations, and you must work out what each operator does from the inputs and outputs shown. Both are popular with employers in technology, engineering, finance and consultancy, because they predict how well you handle unfamiliar problems where there is no prior knowledge to lean on.

A typical question shows you a sequence of frames that changes according to one or more hidden rules, and asks you to select the frame that comes next. The rules are usually simple in isolation (rotation, movement, alternation, addition or removal of elements); the difficulty comes from several rules running at once.

Example question. A square frame contains a small dot in its top-left corner. In each successive frame, the dot moves one corner clockwise. In frame four, which corner is the dot in?

Answer: the bottom-left corner. Frame one: top-left. Frame two: top-right. Frame three: bottom-right. Frame four: bottom-left. Real questions layer two or three rules like this on top of each other, so a reliable technique is to track one element at a time across the sequence rather than trying to absorb every change at once.

Try these tests free, no sign-up needed: Logical Reasoning and Diagrammatic Reasoning

Abstract reasoning tests

Abstract reasoning tests (also called inductive reasoning tests) measure fluid intelligence: your ability to identify the underlying logic in patterns of shapes and figures you have never seen before. Because they use no words or numbers, they are considered less dependent on education or language background, which is one reason employers like them for high-volume graduate recruitment.

Questions typically show a group of figures that share a common rule and ask you which of the answer options follows the same rule, or which figure does not belong. The rules involve properties like the number of sides, shading, symmetry, size and orientation.

Example question. Four figures are shown: a triangle, a square, a pentagon and a circle. Which is the odd one out?

Answer: the circle. The triangle, square and pentagon are all straight-sided polygons; the circle has no straight sides. In a real test the shared property is rarely this obvious, so a useful habit is to run through a mental checklist (sides, shading, size, rotation, count) until one property cleanly separates the group.

If your assessment involves visualising and mentally manipulating objects in three dimensions, see our closely related Spatial Reasoning tests.

Try a free abstract reasoning test, no sign-up needed: Abstract Reasoning

Mechanical reasoning tests

Mechanical reasoning tests measure your understanding of basic physical and mechanical principles: gears, levers, pulleys, springs, pressure and simple electrical circuits. They are standard in recruitment for engineering, manufacturing, maintenance and technician roles, the armed forces and the emergency services.

Each question shows a diagram of a mechanical arrangement and asks you to predict what happens: which way a gear turns, which load is easier to lift, which point moves fastest. You do not need an engineering qualification; the principles are the kind covered in school physics. What the test rewards is applying those principles quickly and without second-guessing.

Example question. Gear A meshes directly with Gear B. If Gear A turns clockwise, which way does Gear B turn?

Answer: anticlockwise. Two gears meshed directly together always rotate in opposite directions. This extends to longer chains: in any line of meshed gears, each gear reverses direction relative to its neighbour, so the first, third and fifth gears all turn one way and the second and fourth turn the other.

Try a free mechanical reasoning test, no sign-up needed: Mechanical Reasoning

Situational judgement tests

Situational judgement tests (SJTs) measure how you weigh up realistic workplace scenarios and choose an appropriate course of action. Rather than testing ability, they assess judgement, professionalism and alignment with the employer's values. They are heavily used by the Civil Service, the NHS, large graduate employers and customer-facing organisations.

Each question describes a scenario, for example a conflict in your team, a mistake you have spotted, or a customer complaint, and offers several possible responses. Depending on the format, you rate each response's effectiveness, rank the responses from best to worst, or pick the best and worst options. The scoring key is usually set by the employer's own senior staff or occupational psychologists, so the "right" answers reflect what that organisation values.

Example question. You notice a colleague has made an error in a report that is due to be sent to a client tomorrow. Your colleague is in meetings all afternoon. What is the best response? (a) Say nothing; it is not your report. (b) Correct the error yourself and send the report without telling anyone. (c) Contact your colleague promptly, explain what you found, and offer to help fix it before the deadline. (d) Report the error straight to your manager without speaking to your colleague first.

Answer: (c). It deals with the problem before it reaches the client, respects your colleague's ownership of the work, and shows initiative without overstepping. Option (a) lets a known error reach a client; (b) fixes the report but hides the issue and bypasses your colleague; (d) escalates before giving your colleague any chance to resolve it, which most scoring keys mark down as a first move.

Try a free situational judgement test, no sign-up needed: Situational Judgement

Personality questionnaires

Personality questionnaires measure your typical behaviour, preferences and working style: how you approach tasks, deal with people, handle pressure and respond to structure. Unlike aptitude tests they are untimed or loosely timed, and there are no right or wrong answers. Employers use them to understand fit with a role and a team, and often to generate interview questions rather than to reject candidates outright.

The usual format presents statements such as "I prefer to plan my work in detail before starting" and asks how strongly you agree or disagree, typically on a five-point scale. Some formats instead force you to choose which of several statements is most and least like you. Your responses are combined into scores on traits (for example conscientiousness, resilience or sociability) and compared with a relevant group of other test-takers.

Example question. "I remain calm when deadlines are tight." Strongly disagree / Disagree / Neutral / Agree / Strongly agree.

How it is interpreted. There is no correct option. Your answer contributes to a trait score, in this case something like composure under pressure, and the employer reads your profile as a whole against the demands of the role. The best strategy is honest consistency: well-built questionnaires ask about the same trait several ways and flag contradictory or implausibly perfect answer patterns, so trying to game the profile tends to do more harm than answering truthfully.

How to prepare for a psychometric test

Preparation is the difference between meeting these tests cold and meeting them on familiar ground. A practical routine looks like this:

1. Find out exactly what you will face. Your invitation email usually names the test publisher (for example SHL, Kenexa or Talent Q) and the test types. Different publishers use noticeably different question styles and time limits, so practise the right format. Our publisher-specific tests and employer-specific tests are built for this.
2. Take a practice test untimed first. Your first attempt at each test type should be about understanding the questions, not the clock. Work slowly, read the explanations for every question, including the ones you got right.
3. Then practise under timed conditions. Time pressure is half the difficulty. Simulate the real thing: timer on, no calculator unless the real test allows one, no interruptions, in one sitting.
4. Review your mistakes properly. Improvement comes from the review, not the attempt. For each wrong answer, identify whether the problem was knowledge (you did not know the method), process (you knew it but misapplied it) or speed (you ran out of time), and target that specific gap.
5. Drill your weakest area, not your favourite one. It is tempting to keep practising the test type you are already good at. Score gains come from the uncomfortable one.
6. Brush up the basics for numerical tests. Percentages, ratios and reading values off charts cover most numerical questions. Ten minutes a day of mental arithmetic sharpens speed noticeably over a couple of weeks.
7. Sort out the practicalities on the day. A quiet room, a reliable internet connection, rough paper, a permitted calculator, and a start time when you are alert. Read the on-screen instructions in full; the number of questions and the timing rules vary between tests, and skimming them is a self-inflicted handicap.

How scoring works: percentiles and norm groups

Psychometric test scores are almost never judged as a simple percentage of correct answers. Your raw score is converted into a percentile by comparing it against a norm group: a reference population of people who have taken the same test, such as previous graduate applicants or current employees in similar roles.

A percentile tells you the share of that group you outperformed. Scoring in the 70th percentile means you did better than 70 per cent of the norm group. It does not mean you got 70 per cent of the questions right; on a hard test, a raw score of half marks can still place you in a high percentile.

Employers typically set a percentile cut-off for progressing to the next stage, and the same raw score can pass at one employer and fail at another depending on the norm group and threshold used. Two practical consequences follow. First, you are being compared with other candidates, not with perfection, so a test that felt difficult may still have gone well. Second, because most candidates cluster around the middle of the distribution, even a modest improvement from practice can move you a long way up the percentile rankings.

Many tests are also negatively affected by wild guessing patterns or, in adaptive tests, become harder as you answer correctly. If your instructions mention negative marking or adaptive difficulty, factor that into your pacing strategy.

Frequently asked questions

What is a psychometric test? A psychometric test is a standardised assessment used to measure mental ability, aptitude or personality in a consistent, comparable way. In recruitment, the most common forms are timed aptitude tests (numerical, verbal, logical, abstract and mechanical reasoning), situational judgement tests and personality questionnaires.

Can I really practise for free without signing up? Yes. Every test type on this page has example questions you can try immediately, free, with no account and no email address required. If you then want full-length tests, worked solutions and performance tracking, you can create an account and see our plans.

Are psychometric tests hard? They are deliberately challenging, mainly because of time pressure rather than the underlying content. Most questions rely on skills you already have; the tests measure how quickly and accurately you apply them. That is exactly why practice helps: familiarity converts time spent decoding the format into time spent answering.

Can you fail a psychometric test? There is no universal pass mark, but employers usually set a cut-off, often expressed as a percentile against a norm group. Score below it and you will not progress to the next stage of that application. The same performance might clear the bar at a different employer.

How long do psychometric tests take? Most individual aptitude tests take between 10 and 45 minutes. If an employer combines several test types in one sitting, allow an hour or more. Personality questionnaires are usually untimed but typically take 20 to 40 minutes to complete thoughtfully.

Do employers check whether you took the test yourself? Increasingly, yes. Many employers ask shortlisted candidates to re-sit a short verification test at the assessment centre or under supervision, and a large mismatch between scores raises a flag. Take the online test yourself, in exam-like conditions, so your verified performance matches.

Should I guess if I am running out of time? Check the instructions first. If there is no negative marking, an educated guess is better than a blank answer. If the test penalises wrong answers, guess only when you can eliminate at least one or two options. In adaptive tests, accuracy usually matters more than raw completion.

Can you cheat on a personality questionnaire? It is unwise to try. Well-designed questionnaires include consistency checks that flag contradictory or implausibly flattering response patterns, and your profile may be explored at interview, where a manufactured persona is hard to sustain. Answer honestly, with your professional behaviour in mind.

How many times should I practise before the real test? Enough that no question format surprises you and your timed scores have stopped climbing steeply. For most candidates that means several timed attempts per relevant test type over one to two weeks, with proper review of mistakes after each attempt, rather than a single cramming session the night before.

Which psychometric test will my employer use? It depends on the role and the test publisher the employer has chosen. Numerical and verbal reasoning are the most widely used for graduate and professional roles; mechanical reasoning appears in engineering and armed forces recruitment; situational judgement tests are common in the public sector and large graduate schemes. Your invitation email normally names the publisher, and you can then practise that specific style via our publisher tests.