AI-generated skills test
Numerical reasoning test
Almost every job now involves a spreadsheet, a dashboard or a metric someone has to interpret correctly, yet a CV line reading "strong analytical skills" tells you nothing about whether a candidate can actually read a chart without misreading it. A Numerical reasoning test replaces that self-description with evidence. It puts figures, tables and short data sets in front of the candidate and asks them to reason to a conclusion, so you observe genuine ability to work with numbers rather than a confident phrase on a resume.
H-Evaluate's Numerical reasoning test focuses on the everyday quantitative work data-driven roles depend on: interpreting proportions and percentages, comparing rates, spotting a trend in a table and rejecting a conclusion the data does not support. Every candidate answers under the same conditions on the same scale, so you can compare a self-taught applicant and a graduate on identical evidence and surface strong reasoners from any background. It maps to the Cognitive pillar of H-Evaluate's five-pillar framework, with items AI-generated for the specific role rather than reused from a static bank. Because it measures a general reasoning ability, it carries its own fairness obligations — best treated as one signal among several and monitored over time.
What it measures
Data interpretation
Reading tables, charts and short data sets accurately — pulling the right figure, comparing series and grasping what a visualisation is and is not saying. The core skill behind any metric-, budget- or dashboard-facing role.
Proportional reasoning
Working confidently with percentages, ratios, rates and changes over time, including percentage-point versus percentage-change distinctions that trip up plausible-looking wrong answers. Central to pricing, forecasting and performance analysis.
Estimation and sense-checking
Judging whether a figure is roughly right before committing to it, and catching an answer that is an order of magnitude off. Signals the instinct to notice when a number in a report simply cannot be true.
Reasoning to valid conclusions
Deciding what the data does and does not support, and resisting an over-reaching inference or a tempting but unsupported claim. This separates careful analysis from a story fitted to the numbers after the fact.
Question formats
Who it's for
Use this test for data-driven roles where reasoning with figures is part of the daily work: analysts, finance and accounting staff, operations and supply-chain roles, product and marketing analysts, consultants, and graduate or early-career hires headed into quantitative work. It is most valuable early in a high-volume funnel, where it screens fairly on the same scale for everyone. Pair it with a Data analysis or Excel test for hands-on tool skill, and with a structured interview to read communication and domain fit that a reasoning score cannot capture.
How to read the results
- 1Read scores in bands relative to a role-relevant comparison group, not as a precise ranking — a band tells you who has cleared a credible bar of numerical reasoning, which is more meaningful than a one-point gap between two candidates.
- 2Match the bar to what the job actually demands. A finance role leans hard on this; a role where numbers appear occasionally should weight it lightly rather than screening people out on a skill the work rarely needs.
- 3Set thresholds in advance and apply them consistently to every candidate, so the standard reflects the role rather than an after-the-fact impression, and pair the result with a structured interview and a hands-on task.
- 4Treat it as one signal among several, and monitor outcomes for adverse impact across groups over time — a skewed ratio is a reason to investigate the test's fit and weighting, not a verdict on any candidate.
AI-generated skills test
Evaluate candidates on this skill with AI-generated questions
Configure a role-tuned assessment and watch it adapt by seniority — no signup.
Related roles
Related reading
Cognitive ability in hiring: use it, don't over-weight it
Cognitive ability in hiring is a strong predictor of potential and the easiest pillar to over-weight. How to use the cognitive signal without an IQ contest.
Do cognitive ability tests reduce bias in hiring?
Do cognitive ability tests reduce bias in hiring? Yes, against gut-feel judgement, but only when calibrated, weighted, and monitored for adverse impact.
Frequently asked questions
What does a numerical reasoning test measure?
A numerical reasoning test measures how well someone interprets data and reasons with figures — reading tables and charts, working with percentages, ratios and rates, spotting trends, and judging which conclusions the numbers actually support. It is not a maths exam of formulas from memory; it assesses the applied quantitative judgement that data-driven roles use daily, giving you objective evidence a CV cannot.
How do you assess numerical reasoning skills?
You give candidates realistic figures — a table, a chart, a short data set — and ask them to reach a conclusion under a sensible time limit, then compare their accuracy against a role-relevant benchmark. With H-Evaluate every applicant answers the same items under the same conditions, so you get a consistent, comparable read on data reasoning rather than an interviewer's impression of who seems analytical.
Are numerical reasoning tests fair for hiring?
They can make a stage fairer by scoring everyone on identical, objective criteria instead of pedigree or presentation, which often surfaces strong reasoners from non-traditional backgrounds. Because it is a general reasoning measure, it also carries a documented adverse-impact risk, so the fairness only holds when the test is job-relevant, weighted as one signal among several, applied at a consistent threshold, and monitored across groups over time.
What is the difference between a numerical reasoning and a data analysis test?
A numerical reasoning test measures the underlying ability to interpret figures and reason to valid conclusions, independent of any particular tool. A data analysis test measures hands-on skill with real data in a working environment — cleaning, aggregating and drawing insight, often in a spreadsheet or query. Reasoning predicts the judgement behind the work; the data test shows whether they can execute it. For most analytical roles the two complement each other.
What score is considered good on a numerical reasoning test?
There is no universal pass mark, because a good score is relative to the role and the comparison group. Read results in bands against candidates applying for similar work rather than as an absolute number, and set the threshold from what the job genuinely demands. A quant-heavy finance role warrants a higher bar than one where figures appear occasionally, so calibrate the standard to the work before deciding who clears it.