Somewhere between your application and your first interview, there is a good chance an employer will ask you to sit a psychometric test. For graduate schemes and high-volume roles it is nearly universal; for experienced hires it is increasingly common. And yet most candidates prepare for the interview in detail while treating the tests as something you either pass or you don't, like an eye exam.
That is a mistake. Psychometric performance is highly trainable. The underlying ability being measured may be stable, but your score on the day depends heavily on familiarity: knowing the question formats, having the core techniques ready, and managing the clock. Candidates who practise for a week routinely outperform stronger candidates who walk in cold. This guide covers what each test type involves and how to prepare efficiently.
Why employers use these tests
Employers use psychometric tests for two main reasons. First, volume: when a graduate scheme receives thousands of applications, tests are a cheap and consistent way to filter. Second, evidence: cognitive ability tests are among the better predictors of job performance, and they are harder to coach in the moment than an interview answer.
Knowing this changes how you should think about them. The test is not trying to trick you. It is trying to rank you against other applicants under identical conditions. Your job is to make sure the score reflects your actual ability rather than your unfamiliarity with the format.
The main test types
Numerical reasoning
You are given data, usually tables, charts, or graphs, and asked questions that require calculation: percentage changes, ratios, currency conversions, proportions, and reading values accurately from a chart. The maths itself is rarely beyond GCSE level. The difficulty comes from time pressure and from multi-step questions where one misread label ruins the answer.
What to practise:
- Percentage change and reverse percentages. These appear constantly. Be able to move between "what is a 12% increase on X" and "X after a 12% increase was Y, what was X" without hesitation.
- Ratios and proportions, including splitting quantities and scaling figures between units.
- Reading charts fast. Train yourself to check axis labels, units, and footnotes before you calculate anything. Most wrong answers come from reading the wrong row, not from bad arithmetic.
- Calculator fluency. Most numerical tests allow a calculator. Practise with the same one you will use, and learn to chain calculations rather than writing down intermediate values.
Verbal reasoning
You read a short passage and judge statements about it as True, False, or Cannot Say. The trap is the third option. "Cannot Say" means the passage does not contain enough information to decide, even if the statement is obviously true in the real world. Candidates fail verbal reasoning by bringing outside knowledge into the test.
The discipline to build: answer only from the passage. If the statement says "the company is the largest in Europe" and the passage says "one of the largest", the answer is Cannot Say, no matter what you happen to know about the company. Practise until this feels natural, because it runs against everyday reading habits.
Logical and abstract reasoning
You are shown a sequence of shapes or patterns and asked what comes next, or which item does not belong. These tests measure fluid reasoning and are the least dependent on prior knowledge, but format familiarity still matters enormously. The patterns draw on a limited library of rules: rotation, reflection, movement around a grid, addition or subtraction of elements, alternation, and changes in size, shading, or count.
Once you have seen a few hundred of these questions, you stop searching blindly and start checking the common rules in order. That systematic approach is the entire advantage practice gives you, and it is significant.
Situational judgement tests (SJTs)
SJTs describe a workplace scenario and ask you to rank or rate possible responses. They measure judgement and alignment with the employer's values rather than raw ability, which is why people wrongly assume you cannot prepare for them.
You can. Two things raise SJT scores:
- Research the employer's values and competency framework. SJT answer keys are written against them. If the organisation publicly prizes collaboration and customer focus, responses reflecting those values will score well. Our guide to researching a company before an interview covers where to find this material.
- Learn the general shape of strong answers. Across most SJTs, good responses involve addressing problems directly but professionally, keeping relevant people informed, taking ownership rather than passing the problem on, and never ignoring an issue in the hope it resolves itself. Extreme responses, either aggressive escalation or complete avoidance, usually rank poorly.
Personality questionnaires
These have no right answers in the way ability tests do, and heavy-handed faking tends to backfire: modern questionnaires include consistency checks, and a profile engineered to look perfect reads as exactly that. Answer honestly but in a work context: describe how you behave professionally, not how you behave on a difficult Sunday.
A one-week practice strategy
You do not need months. A focused week is enough to remove the unfamiliarity penalty.
- Day 1: find out exactly what you are sitting. The invitation email usually names the test provider (SHL, Korn Ferry, Cubiks, cut-e/Aon, and Talent Q are common). Format, timing, and question style vary noticeably between providers, so practise the right one where possible.
- Days 2 to 3: untimed practice. Work through questions slowly and mark every mistake. The goal here is technique, not speed. For every wrong answer, identify whether the cause was method, misreading, or arithmetic, because the fixes are different.
- Days 4 to 5: timed practice. Now practise under real time pressure. Learn your pacing: if the test allows roughly 45 seconds per question, know what 45 seconds feels like, and develop the habit of flagging and moving on rather than sinking three minutes into one stubborn question.
- Day 6: full mock in exam conditions. One sitting, no pauses, phone in another room.
- Day 7: rest and logistics. Check your equipment, browser, and the test deadline. Tired candidates make misreading errors; do not sit the real test at 11pm on the closing day.
On the day
- Do the practice questions the platform offers, even if you feel prepared. They confirm the interface works and settle you into the format.
- Sit the test when you are sharp. Morning suits most people. Avoid squeezing it between meetings.
- Set up like it is an exam. Quiet room, water, rough paper if allowed, calculator ready, notifications off.
- Manage the clock, not just the questions. A flagged question you return to costs less than a rabbit hole. Most tests reward answering more questions adequately over answering fewer perfectly.
- Do not panic about unanswered questions. Many tests are designed so that almost nobody finishes. You are being compared with other candidates facing the same clock.
If the test is part of an assessment centre
Psychometric tests often reappear at assessment centres, either as a fresh sitting or as a verification of an online score, so do not let your practice go stale between stages. Employers routinely re-test on site precisely to check the online result was really you. The tests also sit alongside group exercises, presentations, and interviews, and your energy management across the whole day matters as much as any single exercise. Our guide on what to expect at an assessment centre walks through the full day and how to prepare for each element.
Keep perspective
A psychometric test is one gate, not the whole judgement. Most employers use score bands rather than precise rankings, and once you clear the band, the decision moves to the interview room. So prepare properly, clear the gate, and put the bulk of your energy where the final decision actually gets made. AI Career Mentor can help with that side: practise realistic interview questions, get scored feedback on your answers, and walk into the interview that follows the tests with the same level of preparation you brought to the tests themselves.
Key takeaways
- Test scores are highly trainable: format familiarity and pacing are worth real marks.
- Numerical tests reward percentage fluency and careful chart reading; verbal tests reward answering strictly from the passage.
- Logical reasoning draws on a small library of pattern rules that practice teaches you to check systematically.
- SJTs are preparable: study the employer's values and favour direct, professional, ownership-taking responses.
- A focused week (untimed practice, then timed, then a full mock) removes the unfamiliarity penalty that sinks cold candidates.
