IIF Certified Assessment
Start IQ Test
IIF Certified Assessment
Numerical reasoning is about spotting the relationship that governs a set of numbers — then using it. The arithmetic stays light on purpose. What is being measured is how quickly you find the rule, not how well you were taught to calculate.
The terms look irregular; the gaps do not. They rise by two each time, so the next gap is 10 and the sequence continues at 30. Almost every series item is this move — stop reading the numbers, start reading what happens between them.
Related, but not the same skill. Strong series work with weak data interpretation is a common and perfectly ordinary profile.
| Item type | What it measures | What you do | Example |
|---|---|---|---|
| Number series | Rule induction | Spot the pattern connecting consecutive terms and continue it. | 2, 6, 12, 20, 30 |
| Data interpretation | Reading quantitative displays | Pull the right figures out of a table or chart and compare them. | Which quarter grew fastest? |
| Quantitative comparison | Proportional sense | Judge relative size across fractions, decimals and percentages. | 3/8 vs 0.4 |
| Estimation | Working without exact arithmetic | Get close enough, fast enough, to choose between options. | ~19% of 512 ≈ 97 |
Every item on the real test falls into one of these four types.
Answer one and the sheet shows how the rule was built. Nothing here is scored, and there is no clock on it — the real test is timed.
What comes next?
Choose one
✓ Correct. The gaps are 4, 6, 8 — rising by 2 each time. The next gap is 10, so the answer is 20 + 10 = 30. Reading the gaps rather than the terms is the whole trick.
Not quite — the answer is 30. The gaps are 4, 6, 8 — rising by 2 each time. The next gap is 10, so the answer is 20 + 10 = 30. Reading the gaps rather than the terms is the whole trick.
What comes next?
Choose one
✓ Correct. These are the square numbers: 1, 4, 9, 16 are 1² to 4², so the next is 5² = 25. 36 is the trap — it is 6², one step too far.
Not quite — the answer is 25. These are the square numbers: 1, 4, 9, 16 are 1² to 4², so the next is 5² = 25. 36 is the trap — it is 6², one step too far.
What comes next?
Choose one
✓ Correct. Each term is the sum of the two before it: 2+3=5, 3+5=8, 5+8=13, and so 8+13 = 21. The gaps here are not constant, which is what rules out a simple difference rule.
Not quite — the answer is 21. Each term is the sum of the two before it: 2+3=5, 3+5=8, 5+8=13, and so 8+13 = 21. The gaps here are not constant, which is what rules out a simple difference rule.
What comes next?
Choose one
✓ Correct. Two rules alternate: double, then subtract 2. 7×2=14, 14−2=12, 12×2=24, 24−2=22, so the next step doubles again: 22×2 = 44.
Not quite — the answer is 44. Two rules alternate: double, then subtract 2. 7×2=14, 14−2=12, 12×2=24, 24−2=22, so the next step doubles again: 22×2 = 44.
In which quarter did Product B first outsell Product A?
| Units sold | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| Product A | 120 | 135 | 140 | 138 |
| Product B | 90 | 110 | 145 | 160 |
Choose one
✓ Correct. Compare the two rows quarter by quarter. B trails in Q1 (90 v 120) and Q2 (110 v 135), then overtakes in Q3 (145 v 140) and stays ahead. Nothing here needs calculating — only reading the right pair of cells.
Not quite — the answer is Q3. Compare the two rows quarter by quarter. B trails in Q1 (90 v 120) and Q2 (110 v 135), then overtakes in Q3 (145 v 140) and stays ahead. Nothing here needs calculating — only reading the right pair of cells.
Which of these is the largest?
Choose one
✓ Correct. Put them all in the same form: 3/8 = 0.375, 37% = 0.37, and 0.39 stays as it is. That makes 0.4 the largest. Converting to one common form first is faster and safer than comparing mixed notations in your head.
Not quite — the answer is 0.4. Put them all in the same form: 3/8 = 0.375, 37% = 0.37, and 0.39 stays as it is. That makes 0.4 the largest. Converting to one common form first is faster and safer than comparing mixed notations in your head.
Every sequence above was regenerated from its rule and checked against the printed terms, so each item has exactly one defensible answer.
The distinction is not a reassurance — it changes what the score means.
A maths test asks whether you have been taught a method. It rewards recall of procedures — long division, quadratic formulae, the things that sit in a syllabus. Numerical reasoning asks something else: here is a pattern nobody has explained to you, how fast can you work out the rule?
That is why the arithmetic stays deliberately light. If an item needed a calculator, it would be measuring your calculating, not your reasoning — and the two come apart more often than people expect. Adults who disliked school maths frequently score well here, and people who were fluent at exam arithmetic sometimes do not.
The honest limit runs the same way. A strong score says you read quantitative patterns quickly. It does not say you would enjoy a statistics course, and it is not a substitute for having learned the mathematics itself.
How the scoring works →Fitting for this test: the result page shows its own working rather than hiding the arithmetic behind a number.
How many items you answered correctly. On its own this means nothing — it depends entirely on how hard the items were.
Your raw score is placed against the reference distribution and expressed on the standard scale — mean 100, standard deviation 15.
The share of people who score below you. This is the figure worth quoting, because it needs no explanation of what 116 means.
Illustrative figures shown to demonstrate the layout. Scores use the standard scale, mean 100 and standard deviation 15.
~18 questions · ~15 minutes · instant results
No, and the difference matters. A maths test checks whether you have been taught a method. This checks whether you can find a rule you were never given — the arithmetic involved rarely goes beyond what you would use in a shop. If you needed a calculator, the item would be measuring the wrong thing.
You will not need one, and we would rather you did not. Every item is designed to be settled by reasoning plus light mental arithmetic. Reaching for a calculator usually means you have missed the pattern and started brute-forcing, which costs more time than it saves.
Not necessarily. Numerical reasoning correlates with mathematical attainment but is not the same thing — plenty of people who disliked school maths read patterns and proportions well. If numbers genuinely put you off, the Verbal or Spatial tests measure reasoning without them.
A well-written one should not. Any finite sequence can technically be continued by infinitely many rules, so a fair item is built so that exactly one rule is both simple and consistent with every term shown. Where we could not guarantee that, the item does not ship.
Scores use the standard scale, mean 100 and standard deviation 15, the same convention used across every test on this site. A score of 115 sits around the 84th percentile, 130 around the 98th. See Average IQ by Country for how these figures are usually reported and compared.
Numerical reasoning isolates one ability — finding and applying a rule in numbers, tables and quantities — while a classical test samples several domains and combines them into a single full-scale score. If you want the broader picture rather than one slice of it, the Classical IQ Test is the better starting point; the full set of formats is on the IQ Tests page.
No account or sign-up is required for either the samples on this page or the full test.
Some, but less than you might hope. Familiarity with the format removes easy losses — recognising a ratio pattern quickly, not freezing on a data table — but the underlying ability moves far less than spatial skills do with training. Treat practice as removing noise from your score, not raising your ceiling.
No, and we don’t claim any affiliation with Mensa. Mensa’s own admission tests are administered under supervised, proctored conditions and use their own specific, copyrighted instruments. This test draws on the same general family of numerical and quantitative reasoning tasks that many high-range and admission-style tests use, which makes it a reasonable way to get comfortable with that question format — but a score here is not a Mensa result and can’t be submitted as one.
Take the full profile, switch to words or shapes, or compare all eight formats on the IQ Tests page.
