A numerical reasoning test asks you to take the right figures from charts and tables and work with them to answer business questions. This guide explains what the test measures, walks through a generic worked example, summarises what research says about ability tests in selection, and gives practical advice for candidates and employers.

What a numerical reasoning test measures

Most jobs above entry level involve numbers: reading a sales report, checking a budget, comparing suppliers, or spotting that a trend in a chart is not what it first appears. A numerical reasoning test samples these tasks in a standard form, so that every candidate faces the same data under the same conditions.

The Viewpoint Numerical Reasoning Test has 28 items in 30 minutes. Candidates see 7 sets of data, such as charts, tables and graphs, each followed by 4 questions. The questions cover three skills:

• Reading data: finding the right figures in a chart or table.

• Calculating: addition, subtraction, multiplication, division, percentages and ratios.

• Drawing conclusions: deciding whether a statement about the data is True, False or Cannot tell.

The maths itself is rarely beyond what most people learned at school. What makes the test demanding is choosing the right figures quickly, working accurately under time pressure, and not reading more into the data than it supports.

A worked example

The following is a generic illustration written for this article, not an item from the Viewpoint test.

A table shows a retailer's quarterly sales for two regions. The North sold £240,000 in Quarter 1 and £300,000 in Quarter 2. The South sold £360,000 in Quarter 1 and £405,000 in Quarter 2.

Question 1: By what percentage did North sales grow from Quarter 1 to Quarter 2?

The increase is £300,000 − £240,000 = £60,000. As a share of the starting figure, that is 60,000 ÷ 240,000 = 0.25, or 25%. A common slip is to divide by the later figure instead (60,000 ÷ 300,000 = 20%), which is why answer options often include both.

Question 2: True, False or Cannot tell? "The South grew faster than the North between Quarter 1 and Quarter 2."

South grew by £45,000 on £360,000, which is 12.5%. The North grew by 25%. In cash terms the South's sales are bigger, but its growth rate is lower, so the statement is False. This kind of question rewards people who check what is actually being compared.

Question 3: True, False or Cannot tell? "The North made more profit in Quarter 2 than in Quarter 1."

The table shows sales, not costs, so profit cannot be worked out. The answer is Cannot tell. Many candidates lose marks here by assuming that higher sales must mean higher profit.

What research says about ability tests in selection

Tests of reasoning ability are among the most studied tools in selection. Schmidt and Hunter's influential review of 85 years of research found that general mental ability was one of the strongest single predictors of job performance and training success [1]. Bertua, Anderson and Salgado's meta-analysis of UK studies found similar results, with numerical tests among those predicting both job performance and training outcomes [2].

More recent work has refined these figures. Sackett and colleagues re-examined the corrections used in earlier meta-analyses and concluded that the validity of ability tests had been overestimated in some earlier estimates, while still finding that they predict performance usefully [3]. The balanced reading is that a numerical test gives valuable, objective evidence, but it works best as one part of a wider process.

Using a numerical reasoning test well

For employers, the test is most useful where the role really does involve working with figures, such as finance, analysis, management and many graduate roles. Read results in bands from Very Low to Very High alongside a structured interview, rather than as a pass or fail. A lower score might lead you to ask at interview how the candidate checks their figures, or to look at their work experience with data.

For candidates, the best preparation is practice in the same format. Read the question before searching the data, note the units (thousands, millions, percentages), and keep an eye on the clock so you do not get stuck on one item. A calculator is allowed. Above all, answer Cannot tell when the data does not give you enough information, even if a statement seems likely to be true.

You can try example questions on the Numerical Reasoning Test page, compare our other ability tests, or see our pricing.

Frequently asked questions

How long does the Viewpoint Numerical Reasoning Test take?

It has 28 items and a time limit of 30 minutes.

Can I use a calculator?

Yes. Calculators are allowed as standard, because the test measures reasoning with data rather than mental arithmetic.

What does "Cannot tell" mean?

Choose Cannot tell when the information given is not enough to decide whether a statement is true or false.

Is numerical reasoning the same as numerical analysis?

No. Our Numerical Analysis Test is designed for a different level of role and uses different data. Choose the test that matches the demands of the job.

References

1. Schmidt, F. L. & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology: practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124(2), 262–274.
2. Bertua, C., Anderson, N. & Salgado, J. F. (2005). The predictive validity of cognitive ability tests: a UK meta-analysis. Journal of Occupational and Organizational Psychology, 78(3), 387–409.
3. Sackett, P. R., Zhang, C., Berry, C. M. & Lievens, F. (2022). Revisiting meta-analytic estimates of validity in personnel selection: addressing systematic overcorrection for restriction of range. Journal of Applied Psychology, 107(11), 2040–2068.