Numerical reasoning test for hiring

How the numerical reasoning test holds up against AI help

Each candidate gets a different set of questions drawn from large pools, in a different order, with the answer options shuffled. The browser receives one question at a time, and answer keys never leave HireValid's servers. Timing runs on the server, so reloading the page brings back the same question with only the time that was left, and candidates cannot go back to an earlier question.

Question text cannot be copied or selected, and a blocked copy attempt is recorded in the timeline. Each invitation link is personal and expires. HireValid records signals such as time away from the test tab, pastes into answers and very fast answers, and summarizes them in an Integrity Score with an evidence timeline for the employer to review. It never decides on its own. With the candidate's consent, the employer can also turn on webcam photos.

How to use the result with a work sample and interview

A numerical reasoning result shows whether someone reads and works with numbers accurately. A work sample shows whether they can use those numbers to make a decision. Read the two together, and never use the test result alone.

For shortlisted candidates, share a short table of figures from the kind of work the role does, such as weekly orders, delivery times or staff hours, and ask for three observations and one recommendation, in writing or out loud. Then hold a structured interview with the same questions for everyone. Strong candidates explain how they checked their own numbers, not just what the answer was. The sub-skill results, from data reading to number series, are prompts for what to ask about, not a verdict.

Who should use this numerical reasoning test?

Use this numerical reasoning test when the job involves comparing quantities or reading simple reports: operations staff tracking orders and returns, administrators reconciling figures, sales support teams reading targets, logistics coordinators planning loads, and junior analysts checking a dashboard. The junior data analyst and office administrator roles are typical examples.

It works well as a numerical test for jobs at entry and early-career level, where candidates have little work history to judge, because it uses everyday workplace quantities rather than specialist knowledge. Pair it with a test of the tools the role uses daily, such as a spreadsheet test, so you see both reasoning and practical skill.

For roles where numbers rarely matter, skip it. A numerical aptitude test that has nothing to do with the work adds time for candidates and may screen out people who would do the job well. Match it to real tasks in the job description before you add it. The pre-employment testing guide shows how to pick one test per essential skill.

Is a calculator allowed on the numerical reasoning test?

Yes. An on-screen calculator is available to every candidate, so nobody is helped or held back by how fast they can calculate in their head. That keeps the focus on what the test is built to measure: reading the right figures from a table, choosing the right operation, and handling ratios, percentages and number series correctly.

The calculator does not answer the question for anyone. A candidate still has to decide which values matter, which figure is the starting point for a percentage, and what the rule behind a series is. So the result reflects reasoning with numbers, not calculation speed.

The wording is kept plain and the quantities are everyday ones, such as hours, units, distances and visitors, in metric units, with no money symbols or local knowledge. If a candidate needs extra time as an accommodation, say in the invitation how to request it before they start.

What a data interpretation test catches

Many number mistakes at work are not calculation errors. They come from reading the wrong figure, using the wrong starting point or mixing up periods. A data interpretation test draws on those habits, so it is useful even for roles that never touch advanced math.

Three common errors:

• The wrong base for a percentage. Visitors falling from 500 to 450 is a fall of one tenth measured from 500, not one ninth measured from 450.
• Relative change versus percentage points. A return rate moving from 5 to 8 returns in every hundred orders is a rise of 3 percentage points, or three fifths in relative terms. Mixing the two leads to wrong conclusions in a report.
• Ignoring units and periods. Comparing a weekly figure with a monthly one produces confident but wrong answers.

The data reading and ratios and percentages questions in this test call on the same care: finding the right value and the right starting point before calculating. In the interview, ask candidates how they check those steps in their own work. Listen for whether they can explain why the base figure matters.