Skip to main content

Understanding Thrive assessments

Please read this before you complete your certification. It takes a few minutes, and everything the quiz covers is in here.

S
Written by Software Tech

Before you start building campaigns with Thrive, it helps to understand how our assessments work and why they predict job performance so well.

This guide walks through the science behind the questions, the different assessment types, and how to get the most out of them.

What goes into creating an assessment

Every question is built by world-class psychologists, then crafted, tested and validated. Four things stand behind the accuracy of your results:

  1. Building the assessments. Writing questions is only the starting point. Our psychologists define what each question should measure, trial it with a large, representative sample of people, and check it for reliability and bias. We also check that scores are linked to real job performance. A question doesn't become accurate just because it sounds plausible.

  2. Reviewing the questions. Looking through individual questions won't tell you whether an assessment is accurate. That review is already part of how we build every test. What's useful for you is knowing what the results mean and what to explore with the candidate.

  3. Population benchmark. Every score is compared against Thrive's working population benchmark. It shows how a candidate compares with a large group of working adults, not just with others in the same campaign.

  4. Keeping assessments fair. Fairness is reviewed monthly. We combine expert review of question wording with statistical checks, looking at whether particular groups consistently score higher or lower on individual questions. Questions found to introduce unfair disadvantage are removed and replaced, and this repeated process improves fairness over time.


The three assessment types

Thrive uses three assessment types. Two of them, cognitive ability and personality, deliberately do not contain job-specific content. The third, skills, is built around the job itself.

Cognitive ability

Cognitive ability is the single most accurate predictor of job performance, ahead of interviews, experience and qualifications. Decades of research consistently show it outperforms every other assessment type.

There are four cognitive assessments:

  • Numerical Reasoning. Interpreting data, tables and charts to draw accurate conclusions.

  • Verbal Reasoning. Understanding written passages, drawing inferences and telling facts from assumptions.

  • Logical Thinking. Pure pattern recognition using abstract shapes, with no language or numbers. This makes it the fairest measure across culture, language and education.

  • Checking. Speed and accuracy of information processing under time pressure.


Personality

Personality questions measure underlying psychological traits, not job-specific situations. The behaviours and traits we look for can be specific to a role, but the questions themselves measure those underlying traits rather than job knowledge. That is exactly what makes the results more accurate and less biased.

Personality is built on the Big Five / OCEAN model, using a forced-choice format that prevents candidates scoring high on everything and produces a true hierarchy of behaviours. There are no right or wrong answers, which removes coaching and faking.


Skills

Unlike cognitive and personality assessments, skills questions are directly related to the job. They test the specific knowledge and competencies a role needs on day one, and they adapt to each role, whether that is a developer, marketer or finance hire.


Questions we hear all the time

Great questions come up often. Here is what the science says:

Interviews

  • I'm a good interviewer. Isn't that enough? Experience is valuable, but even experienced interviewers see only a small sample of behaviour, and rapport and confidence can sway their impressions. Use Thrive's results to plan structured questions and score the evidence consistently.

  • What if a score doesn't match my interview impression? People adapt their behaviour to situations. Someone may be outgoing in an interview without preferring frequent social interaction. Check what the score measures and ask for examples of how they behave in different situations.

    Assessment questions

  • Can I add my own questions? Custom questions, whether written in-house or by AI, haven't been tested or validated, so they don't reliably predict job performance. Validated questions are what give the results their accuracy.

  • Should I reject a candidate below the preferred range on a trait? Not automatically. One trait is only part of their profile. Consider what it means for the role and their wider profile, then explore it with structured interview questions.

    Candidate results

  • A candidate disagrees with their personality results. What do I say? Their results come from the answers they gave, scored and benchmarked against the working population. Someone may see themselves as highly organised compared with people they know, yet score differently against the benchmark. Ask which findings they disagree with and explain what the scores mean.


Understanding scores

Personality scores. Personality scores are shown as percentiles from 1 to 99, compared against Thrive's global working population benchmark. A score of 40 means the candidate scored higher than 40% of that group. It reflects how strongly they identify with the trait at work, not their ability, and it isn't a pass or fail.

Cognitive scores. Cognitive raw scores are converted into percentiles from 1 to 99, compared against Thrive's continuously updated working population benchmark. A score of 55 means the candidate performed as well as or better than 55% of that group. It isn't the percentage of questions answered correctly, and it falls within the Moderate band.

The overall score. The overall score isn't an average of the individual results. Thrive combines the individual results and compares them with the working population benchmark. Scoring well across many traits is rarer than scoring well on one, so a candidate with strong scores across several traits can get a higher overall score than one who scored highly on a single trait. It also works the other way: consistently lower results can produce a lower overall score, even if one trait is very high.


Assessments always come before interviews

Assessments belong early, before interviews, at the application stage. Using them before interviews helps you make better-informed shortlisting decisions, reduce unnecessary interviews and spot strengths that might not stand out on a CV.

The candidate's results and Thrive's suggested interview questions then help you structure every interview. This gives you comparable evidence for each candidate, so decisions depend less on CV presentation or rapport and are easier to explain.


Our assessments predict real job performance

We track assessment scores against real job performance to check our science holds up. Across 800+ customers worldwide and millions of candidates assessed, there is a clear positive correlation between how people score on Thrive and how they go on to perform in their roles.

Testing on your own team: what best practice says

Trying it yourself. Trying an assessment can show you what candidates experience, but whether you agree with your own result doesn't tell you whether it's accurate. Accuracy comes from reliability (whether it measures consistently) and validity (whether scores mean what they claim and predict performance).

When a strong employee scores low. An assessment measures particular attributes, not someone's entire contribution or accumulated expertise. An experienced employee may have developed effective ways of working that compensate for a relative weakness. Their motivation and approach during testing may also differ from a candidate's. A single employee's result can't establish whether an assessment predicts performance across a wider population.

Testing your top performers. A small group can produce apparent patterns by chance. Selecting only top performers also gives no comparison with people performing at other levels, so it can't show which attributes distinguish performance. Different combinations of strengths can lead to success, and copying your current team risks reproducing previous hiring preferences. To decide which attributes matter, use a structured analysis of what the role actually requires.


Supporting candidates with a disability or neurodivergence

If a candidate says their disability or neurodivergence may affect how they complete the assessment, start with the barriers they anticipate and the support they need. People with the same condition may need different adjustments, so avoid assuming one approach fits everyone.

Where extra time is appropriate, you can add it directly in the Thrive platform. Contact Thrive if the candidate needs further support or an alternative before they proceed.


Things to get right

  1. Use assessments early. Assess at the application stage so every interview is focused, fair and backed by data.

  2. Assess your employees on our Employee Development platform. The employee tool is included in your subscription and uses competency and wellbeing data to make hiring decisions more accurate.

  3. Don't test validity yourself. Trust the science and review our published research and validation studies.

  4. Make adjustments where needed. Ask candidates about any barriers they expect, and add extra time in the platform where appropriate. This works well if you want accessibility to be visible early, but the new Fairness section will already cover it.


Once you have read through this, you are ready to complete your certification.

Did this answer your question?