Assessivo

Methodology

How the Assessivo IQ Test Works

What the test measures, how each test is put together, exactly how your score is calculated, and what it can and can't tell you.

Last updated: October 13, 2026

In short: each test gives you 36 questions, 9 in each of four areas of reasoning, chosen from a pool of 960. Your score depends on how many you answer correctly: half right gives 100, and each extra correct answer adds 2.5 points. The percentile is read from the standard IQ bell curve. Scores are estimates on a provisional scale, and we explain below how we are calibrating it against real results.

What the test measures

The test covers four kinds of reasoning, all closely linked to general intelligence, the broad ability that IQ tests are designed to estimate. Our guide What is IQ? explains these abilities in more depth.

  • Fluid reasoning: finding the rule in unfamiliar patterns, such as 3×3 pattern grids, figure sequences and odd-one-out shapes.
  • Logical reasoning: drawing valid conclusions from statements, including syllogisms, ordering and seating puzzles, analogies and odd-one-out words.
  • Spatial reasoning: picturing shapes as they turn, flip and stack: rotations, mirror images and counting cubes.
  • Numerical reasoning: number series and short word problems on percentages, rates, ages, costs and averages.

These are the same broad families of task used in professional tests such as the Wechsler scales, although our questions, format and scoring are our own. You can try worked examples of each type in our sample IQ test questions.

The question pool

960questions in the pool
4areas, with about 240 questions each
34different question types
5difficulty levels, from easy to hardest

Every question is multiple choice with four options and one correct answer, and every question has a written explanation, which appears in the optional detailed report. Questions are generated from templates, one for each question type, and each is assigned a difficulty level from 1 (easy) to 5 (hardest) when it is created. Each question has a unique fingerprint, so the same puzzle can't appear in the pool twice.

How each test is put together

Every time you start, a new set of 36 questions is drawn from the pool, following the same blueprint so that every test has the same structure:

  • 9 questions per area, so no area counts for more than another.
  • The same difficulty mix in every area:
  • Easier questions first. Within each area the questions run from easiest to hardest, and the four areas take turns, so the test gets steadily harder as you go.
  • Variety. No single question type can dominate. For example, a test has at most six pattern grids and at most five rotation questions.
  • Fresh questions on retakes. Your browser remembers the last 108 questions it was shown, and we pick questions you haven't seen first. Clearing your cookies resets this.

The time limit

You have 25 minutes for the 36 questions, about 42 seconds each. A timer shows the time left, and when it reaches zero your answers are submitted automatically. Unanswered questions count as incorrect, and there is no extra penalty for a wrong answer, so if you are running short of time, a reasoned guess is better than leaving a question blank.

The limit is generous compared with employer tests, which are designed so that almost nobody finishes. Our aim is to measure how well you reason, more than how fast.

How your score is calculated

Your score depends on one thing: how many of the 36 questions you answer correctly. Each correct answer counts the same, whatever its difficulty. The share you get right is placed on the familiar IQ scale with this formula:

IQ score = 100 + (share correct − 0.5) × 90

So getting exactly half right (18 of 36) gives 100, every extra correct answer adds 2.5 points, and every one fewer takes 2.5 away. The lowest possible score is 55 (none right) and the highest is 145 (all 36 right). Try it:

Score calculator

Estimated IQ score115
Percentile84th
Likely range103–127

This uses the same formula as the test itself. The likely range is explained in How precise is the score?

To see what each level means, explore our IQ scale and score ranges, or look up any score from 55 to 180 on its own page, for example an IQ of 115.

What your percentile means

Your percentile shows where your score sits on the standard IQ bell curve, which has an average of 100 and a standard deviation of 15. A score of 115, for example, is at the 84th percentile: on that curve, about 84% of people would score lower.

It is important to be clear about what this is. The percentile comes from the theoretical bell curve, not from ranking you against the other people who have taken Assessivo. It tells you where your score would fall if the test's scale matched the general population exactly. Making that true is the purpose of the calibration work described below. Percentiles are shown between the 1st and 99th.

Your four area scores

Alongside your overall score, you get a profile of the four areas. Each area score is simply the percentage of that area's nine questions you answered correctly. With only nine questions per area, a single answer moves an area score by 11 points, so treat the profile as a guide to your relative strengths, not as four separate IQ scores.

The optional detailed report adds more: every question with its explanation, how your accuracy changed as the questions got harder, how long you spent in each area, and a seven-day practice plan for the area with the most room to grow.

How precise is the score?

No test measures anyone perfectly, and a short test less so than a long one. If you took a different set of 36 questions of the same difficulty, chance alone would move your score a little, because some questions suit you better than others and a few answers are lucky or unlucky guesses.

The detailed report shows this as a likely range. We work it out from the number of questions and the share you got right, using the standard statistical margin for a proportion, and show the range that would cover your score about 90% of the time. For a score of 100 it is roughly 88 to 112; it is narrower near the top and bottom of the scale. Professional tests such as the WAIS are more precise mainly because they are much longer and are given one-to-one by a trained psychologist.

Two more things can move a score between attempts:

  • Practice. Scores usually rise on a second attempt as the question types become familiar, even with new questions. Your first attempt is the best estimate of your ability.
  • Conditions. Tiredness, distractions, a small phone screen or a poor connection can all lower a score. Take the test somewhere quiet, when you are fresh.

Our article Are online IQ tests accurate? covers these limits in more detail and includes a checklist for judging any online test.

Calibration: what we're working on

An IQ scale is meant to be normed: built from the results of a large group of real test takers, so that 100 is the typical score and the percentiles describe real people. Professional tests are normed on thousands of carefully selected people. Our scale is currently provisional. It assumes that answering half the questions correctly corresponds to an average score, and it doesn't yet use real results to check that assumption.

NowProvisional scale. Scores and percentiles come from the formula above. The detailed report compares your score with the real average of everyone who has finished the test, once at least 50 people have.
NextCalibration from real results. When enough tests have been completed, we will use the anonymous answer data to set 100 at the real typical score, measure how hard each question actually is instead of relying on its assigned level, weight harder questions accordingly, and base percentiles on how people really score.

Online test takers are a self-selected group, not a representative sample of the population, so even a calibrated online scale won't match a clinical norm exactly. When we change the scoring, we will update this page and say what changed and when.

See the method in action36 questions, 25 minutes, and your score, percentile and area profile straight away.
Take the free IQ test

What the test is not

  • Not a clinical or diagnostic test. It is not given by a psychologist, and it can't diagnose a learning difficulty, giftedness or any other condition.
  • Not adjusted for age. Everyone's score is calculated in the same way. Professional tests compare you with people of your own age, which matters most for children and older adults. Assessivo is for people aged 13 and over.
  • Not for decisions. Results should not be used for medical, educational or employment decisions. If you need a formal IQ assessment, for example for school placement or a clinical question, contact a qualified psychologist.
  • Not a measure of everything. Reasoning tests leave out creativity, knowledge, practical skills and much else that matters. Our guide What is IQ? explains what IQ predicts and what it misses.

The data we use

You can take the test without an account, a name or an email address. For each completed test we store your answers, whether each was correct and how long you spent on each question. We use this to show your results and report, and in anonymous, aggregated form to improve the questions and calibrate the scale. Our privacy policy explains how data is handled, and our editorial policy covers how we research and write our articles. You can read more about who runs Assessivo on our about page.

Sources