What Is a Psychometric Assessment? Types and How to Prepare | Cohesyve

A psychometric assessment is a standardised test of your abilities or your personality. Everyone takes the same questions under the same rules and is scored the same way, so an employer can compare applicants on the same measure. "Psychometric" simply means psychological measurement. You'll usually meet one early in an application, online.

Two kinds of test: with and without right answers

The test publisher Hogrefe divides psychometric assessments into two families:

Maximum performance tests, which have right and wrong answers. Aptitude and cognitive ability tests are in this group. Your job is to get as many right as you can, usually against the clock.
Typical performance measures, which have no right answers. Personality and motivation questionnaires describe how you tend to behave and what you prefer.

Situational judgement tests sit between the two: they ask about realistic work situations, and some answers are better than others.

Aptitude tests: numerical, verbal and logical reasoning

Aptitude tests are usually multiple choice and timed, and some are designed so that almost nobody finishes. Criteria says fewer than 1% of people answer all 50 questions of its CCAT in the 15 minutes allowed. The main types follow, each with an example question written for this guide.

Numerical reasoning

Questions about figures in tables and charts: percentages, ratios and changes over time. The arithmetic is usually simple; the skill is finding the right figure quickly.

Example. A shop sold 12,000 units in January and 15,000 in February. By what percentage did sales rise?

A) 20%
B) 25%
C) 30%
D) 3%

Answer: B. The rise is 3,000, and 3,000 ÷ 12,000 = 0.25, or 25%. Dividing by February's figure instead gives 20%, a common trap.

Verbal reasoning

A short passage, then statements to judge as True, False or Cannot say, using only what the passage says.

Example. "The support team moved to a four-day week in March. Since then, response times have fallen and sickness absence has dropped. The company hasn't decided whether to extend the trial to other teams." Statement: The four-day week caused response times to fall.

Answer: Cannot say. The passage says response times fell after the change, not because of it.

Logical reasoning

Inductive reasoning, also called abstract or diagrammatic reasoning: spot the rule in a sequence, usually of shapes, and choose what comes next.
Deductive reasoning: decide what must follow from a set of statements.

Example (inductive). What comes next: 3, 6, 11, 18, 27, ...?

Answer: 38. The gaps grow by two each time: 3, 5, 7, 9, then 11.

Example (deductive). All the engineers in this office have a security pass. Sam works in this office and has a security pass. Does it follow that Sam is an engineer?

Answer: No. Every engineer has a pass, but nothing says that only engineers do.

Other aptitude tests check spatial reasoning (turning shapes in your head), mechanical understanding or how accurately you check information. Some employers send a general test that mixes types at speed, such as the CCAT (verbal, maths and logic, and spatial questions) or the Predictive Index's cognitive assessment (verbal, numerical and abstract reasoning).

Cohesyve

Get used to being assessed before it counts

Run an AI job simulation of the real work and get a scored report on where you stand. 3 free tests, no card.

Start practising free →

Personality questionnaires

These ask how you usually behave or what you prefer. They're usually untimed, and there are no right or wrong answers. Two formats are common:

Rating statements, for example from Strongly disagree to Strongly agree: I like to plan my work in detail before I start.
Forced choice: from a group of statements, you pick the one most like you and the one least like you: I enjoy meeting new people. / I finish what I start. / I stay calm under pressure.

Diagram of two personality question formats: a Strongly disagree to Strongly agree rating scale, and forced choice of most and least like you

Your answers become scores on traits such as how sociable, organised or assertive you are, compared with a norm group (see below) and sometimes with a profile of what the job needs. A higher score isn't a better one: the Predictive Index says exactly that about its behavioural assessment.

Forced choice is also meant to make faking harder: Talogy says the forced-choice questions that make up most of its Caliper Profile help it resist deliberate manipulation.

Situational judgement tests

A situational judgement test (SJT) describes a realistic work problem and asks you to judge some possible responses: pick the best and worst, rank them or rate each one. The US Office of Personnel Management explains that answers are scored against job experts' judgements of the best and worst options.

Example. A customer emails to complain that their order arrived late. You can see the courier caused the delay. How effective is each response?

a) Explain that the courier was at fault and send the courier's contact details.
b) Apologise, explain what happened and say what you'll do to stop it happening again.
c) Offer a full refund straight away, without checking the order.
d) Forward the email to your manager.

Answer: (b) is the most effective. It takes responsibility for the customer's experience without blaming anyone or overstepping your authority. (a) passes the problem on, (c) may offer more than you're allowed to, and (d) hands off something you could handle.

Other assessments you might meet

Motivation and values questionnaires ask what kind of work you prefer. Wonderlic's assessment includes one, with 58 items.
Game-based assessments measure abilities and traits through short games rather than questions.
Emotional intelligence and integrity questionnaires appear in some processes too.

Which test have you been sent?

The invitation often names the publisher. Our guides cover several common ones in detail:

SHL: numerical, verbal, inductive and deductive reasoning tests, and SJTs. See our guide to SHL tests.
Wonderlic: a 12-minute cognitive section of 50 questions, plus untimed motivation and personality sections; about 35 minutes in all. See the Wonderlic test.
CCAT: Criteria's 50 questions in 15 minutes, with no calculator. See the CCAT.
Predictive Index: a cognitive assessment of 50 questions in 12 minutes, and an untimed behavioural assessment that takes about six minutes. See the Predictive Index test.
Caliper: 98 untimed questions, mostly forced choice, with some abstract reasoning; most people take 60 to 75 minutes. See the Caliper test.
Situational judgement tests from any publisher: see our guide to SJTs.

How you're scored: norm groups and percentiles

Raw score. On an aptitude test, the number of questions you got right.
Norm group. The people your score is compared with: a large group who have taken the same test, chosen to represent the people it's meant for, such as graduates or experienced managers.
Percentile. The percentage of the norm group you scored higher than. The 70th percentile means you beat 70% of that group, not that you got 70% of the questions right. Criteria, for example, sends employers your CCAT raw score and your percentile.

The same raw score can be a high percentile against one group and an average one against another, so it matters who you're compared with. Personality results are compared with a norm group too, but as a description, not a grade.

Flow diagram: a raw score from a ticked checklist is compared with a norm group of people to give a percentile on a bell curve

Employers use results in different ways. Some set a minimum for the next stage; others look at the whole profile, or use it to choose interview questions. Some aptitude tests are adaptive, giving harder or easier questions depending on your answers, so finding a test hard doesn't mean you're doing badly.

How to prepare

Use the publisher's own practice. SHL Direct has practice versions of SHL's reasoning tests, SJT and personality questionnaire, Criteria offers free practice tests, and the Predictive Index publishes sample questions. SHL warns that its practice tests may not match the difficulty of the real thing, so use them to learn the format.

For aptitude tests, practise against the clock. Work out how long you can spend on each question, and move on from any that would take too long. On a test scored by the number you get right, a quick, easy question is worth as much as a slow, hard one.

Brush up on the basics. Percentages, ratios and reading charts for numerical tests; answering only from the passage for verbal ones. Check whether you can use a calculator: the Wonderlic and the CCAT don't allow one.

For personality questionnaires, answer as yourself. Go with your first reaction and be consistent. Don't play an imagined ideal candidate: forced-choice questionnaires are built to resist it, and if it worked, you could end up in a job that doesn't suit you.

For SJTs, learn what the organisation values, and answer with what you should do, not what's easiest.

Set up properly. A quiet room, a laptop rather than a phone if there are charts to read, a stable connection and enough time to finish in one sitting.

Ask for adjustments early. If you have a disability, ask the employer about adjustments, such as extra time, before you start. In the UK, employers must make reasonable adjustments, including to their recruitment process, and US employers have a similar duty under the ADA, according to the EEOC.

Ask what happens to your results. Hogrefe's guidance says you should be told why a test is being used, how the results will be used, who will see them and what feedback you'll get. If nobody has said, ask.

Where Cohesyve fits

Psychometric tests are usually one stage of several, and many employers also ask you to do a task from the job itself. Cohesyve, which publishes this guide, runs practice job simulations for your role, with a scored report on where you lost marks. It isn't a psychometric test; for those, the publishers' own practice tests are the best preparation.

Sources

Checked 3 October 2026.