Numerical questions ask you to draw conclusions from data such as tables, charts, and percentages. Most people lose marks not on the arithmetic but on a few predictable traps: confusing percentage change with percentage points, misreading ratios, and letting the first number they see anchor their answer. Knowing these traps is the best defense.
Whether you are sitting these questions for a frontline role or a senior appointment, the same small set of mistakes accounts for most lost marks. I have watched capable people get the arithmetic right and the answer wrong, because the trap was in how the numbers were framed, not in the calculation. This article sets out a simple way to approach a numerical question, then names the specific traps that catch people. It does not list every question format or how to prepare, which are covered separately. It is about the few seconds where a sound calculation still leads to the wrong answer.
What kinds of numerical questions will you face?
Numerical questions fall into a few recurring shapes. The most common is data interpretation, where you read a table or chart and answer questions about it. Around that sit percentages and percentage change, ratios and proportions, rates such as speed or cost per unit, currency and unit conversions, and trends across time.
The arithmetic behind these is rarely the hard part. The difficulty is choosing the right operation, reading the correct figure from a busy table, and keeping units consistent. Most wrong answers come from one of those three slips, not from a failure to calculate. That is why the method you use matters more than your raw speed with numbers.
How should you approach a numerical question?
Approach each question in a fixed order. Read the question before you study the data, so you know exactly which figures you need. Find those figures in the table or chart and check their units. Estimate the answer roughly in your head. Then calculate, and check that your answer is close to your estimate.
The estimate step is the one most people skip, and it is the most valuable. A rough estimate catches the large errors, the answer that is ten times too big or has the wrong sign, which are the errors that cost whole marks. If your calculated answer is far from your estimate, you have made a slip and should look again before moving on. Work tidily, one step at a time, and resist the urge to rush the reading in order to start calculating.
What is the trap with percentages?
Percentages are the single richest source of traps in numerical questions, because everyday habits of speech do not match the precise meaning the questions require. The most common trap is confusing a percentage change with percentage points.
If a rate rises from 20 percent to 25 percent, that is a rise of 5 percentage points, but a rise of 25 percent in relative terms, because 5 is a quarter of 20. The question may ask for either, and the wrong reading gives the wrong answer. A related trap is relative against absolute change. Research on understanding risks shows that the same figure can look far larger when it is framed as a relative change than as an absolute change. People are misled by the framing even when the underlying numbers are identical. When you meet a percentage question, pause and ask precisely which quantity the percentage is being taken of. That single habit prevents most percentage errors.
What is the difference between percentage and percentage points?
A percentage point is the simple arithmetic difference between two percentages, while a percentage change expresses that difference relative to the starting value. Moving from 20 percent to 25 percent is 5 percentage points, and also a 25 percent increase relative to the original 20. They are different numbers and answer different questions, so read the question carefully to see which one is being asked. Mixing them up is one of the most frequent mistakes in numerical reasoning.
What is the trap with ratios and proportions?
Ratios and proportions are the second great source of error, because people tend to focus on the top number of a fraction and neglect the bottom one. This makes a larger raw count look like a better chance even when the proportion is worse.
Research on numeracy calls this denominator neglect. People shown two options often prefer 9 chances in 100 over 1 chance in 10, because 9 feels bigger than 1, even though 1 in 10 is the better proportion. In a numerical question, this surfaces when you compare ratios with different denominators, or read a proportion straight from a table without converting it to a common base. The guard is to convert ratios to the same denominator, or to a percentage, before you compare them. Never compare the top numbers alone.
Why do people get ratio questions wrong?
They get them wrong because they compare the numerators and ignore the denominators, so the option with the larger raw count looks better even when its proportion is smaller. The fix is mechanical. Put the ratios on a common base, by converting each to a percentage or to a fraction with the same denominator, and only then compare. A ratio means nothing until you know what it is a ratio of, so always anchor it to its base before you judge it.
How do anchoring and estimation affect your answers?
The order in which numbers appear can quietly bias your answer. The first figure you see, or a prominent number in the question, can pull your estimate toward it, even when it is not the relevant figure.
This tendency is called anchoring, and it is one of the most reliable findings in the study of judgment. People asked to estimate a quantity stay close to whatever number was put in front of them first, adjusting too little away from it. In a numerical question, the anchor might be a large total at the top of a table, or a striking figure in the stem, with no bearing on what is being asked. The defense is to form your own rough estimate from the relevant figures before you look at the answer options, so that a distracting number cannot pull you off course. Trust the estimate you built, not the number that happened to catch your eye.
Key takeaways
1. Numerical questions are usually lost on reading and framing slips, not on the arithmetic itself.
2. Approach each question in order: read the question first, find and check the figures, estimate, then calculate and compare against the estimate.
3. The estimate step catches the large errors that cost whole marks, so never skip it.
4. Distinguish a percentage change from percentage points, and always ask which quantity a percentage is being taken of.
5. For ratios, put the figures on a common base before comparing, because focusing on the top number and neglecting the bottom is a classic error.
6. Beware anchoring. Form your own estimate from the relevant figures before a prominent but irrelevant number pulls you off course.
7. A tidy, step by step method beats raw speed, because the traps reward care, not haste.
What this means for you
You do not need to be fast with numbers to do well. You need a method and the discipline to follow it. Read the question first. Estimate before you calculate. Treat every percentage and every ratio as a question about what the number is being measured against. And build your own estimate before you trust any figure the question puts in front of you.
These habits are worth rehearsing, because under time pressure you will fall back on whatever you have practiced. They sit inside the wider set of assessments that employers use to make a fair selection judgment. Reasoning carefully with numbers is simply showing that part of your thinking at its best, and it serves you well beyond any test.
Related reading on The Human Capital Hub
For the wider picture of how these assessments work, read our psychometric tests guide. For a longer view of why reasoning with numbers is studied so closely, our article on cognitive ability and broader social outcomes offers useful context.
