The Mensa Norway IQ test explained: format, scores, percentiles and real admission

The Mensa Norway online test: format, timing and scoring

Mensa Norway hosts the free online test most people mean when they say they "took the Mensa test". Its page describes the format plainly: 35 puzzles, each a three-by-three grid of visual patterns with one box empty and six alternatives to complete it, to be solved within a 25-minute limit. No registration, no fee and no specialised or mathematical knowledge are required; you choose an age range (16–17, 18–50, 51–60 or 61+) and start.

• Every puzzle is worth one point and the puzzles get progressively harder.
• There is no bonus for finishing early, so use the whole 25 minutes.
• There is no penalty for a wrong answer, so guess whenever you are unsure.
• The result is an indicative score between 85 and 145, where 100 is the population average.
• The page itself says the test is not a substitute for professional intelligence tests such as those administered by Mensa and licensed psychologists.

That last sentence is the one to keep. The Norway test is a well-made practice run in the matrix format, taken unsupervised on whatever screen you happen to have; a score from it is an indication, never a Mensa qualification.

What your Mensa Norway score means

The Norway test prints one number between 85 and 145, where 100 is the population average, and its result screen states the scale behind it: a standard deviation of 15. On that scale a score fixes a percentile, the share of people at or below it, because the percentile is a property of the normal curve rather than of anyone's answers. The table applies that curve, with a mean of 100 and a standard deviation of 15, to every fifth score the test can report. That is an assumption taken from the test's own statement of its scale. The page does not describe the sample its scores were set on, so the table is the arithmetic of the scale and not a measurement of the people who take the test.

145 is the ceiling. It is the highest score the test reports, so a result of 145 means 145 or higher: about the 99.9th percentile, with 1 in 741 people or fewer scoring higher, and nothing beyond it. Few puzzles in any test are hard enough to separate people at that end of the curve, which is why a number near the top is the least precise one in the table. 85 is the floor in the same way, and means 85 or lower.

Before you start you choose an age range: 16–17, 18–50, 51–60 or 61+. The page does not say how that choice changes the score, so the table applies the same curve to every age range. If you have a Norway result in front of you, the percentile printed beside it is the test's own figure; the row for your score here should agree with it.

If you have not taken it yet, the official test is at test.mensa.no. Take it once and keep that first score: a second attempt on the same 35 puzzles partly measures how well you remember them.

For a second opinion on different puzzles, our free IQ test reports how many you solved by question type and gives an IQ-style practice estimate. It is not on Mensa Norway's scale and is never turned into a percentile, so its range does not belong in the table above.

Why your score differs between tests

A different number from another test is ordinary, and it rarely means that either test is wrong. Each test is its own scale, and six things move the number without any change in the ability it measures:

• Different spreads. Most current tests use a standard deviation of 15 and some do not: American Mensa's qualifying list gives 130 on the Stanford-Binet 5, 132 on the earlier Stanford-Binet and 148 on the Cattell for the same top-2% requirement. A Norway 130 and a Cattell 148 can describe the same standing.
• Different reference groups. Each test is scaled on its own sample, in its own decade. Measured performance on the same old items has drifted upward by roughly 10 to 20 points of population IQ every thirty years or so, the pattern called the Flynn effect, so a test whose norms are decades old compares you with a group that would score lower today.
• Measurement error. Every score is an estimate with a band of several points around it, so five or six points between two good tests is the same result read twice.
• Conditions. The Norway test runs unsupervised, on whatever screen you have and with whatever interruptions; a supervised test controls all of that.
• Practice. A familiar item format, or the same puzzles a second time, can raise a score without any change in the ability it is meant to measure.
• Ceilings. Norway stops at 145. A test that reaches 160 can print a number Norway cannot, without measuring anything different.

So compare percentiles rather than raw numbers, and only between tests whose scale you know. A gap of twenty points usually says more about the scale, the conditions or a repeated attempt than about you.

Is the Mensa Norway test official?

Yes and no. It is Mensa Norway's own test, on Mensa Norway's own site, so the puzzles and the score are genuinely theirs; a copy elsewhere is not. But it is an online indication, not an admission test: its page says it is no substitute for the professional intelligence tests that Mensa and licensed psychologists administer, and a result from it is not accepted for membership anywhere.

Practice tests elsewhere, ours included, are unofficial. The Mensa Norway practice test on MyTestAtlas uses original puzzles written in the same format, 35 three-by-three grids with six options against a 25-minute clock, and none of Mensa Norway's own puzzles. MyTestAtlas is not affiliated with or endorsed by Mensa.

Why a practice site writes its own items rather than copying a published test is explained in which psychological tests may be reproduced.

How hard it is, and how to pace it

Twenty-five minutes for 35 puzzles is about 43 seconds a puzzle on average, but the time is not spent evenly. The puzzles get progressively harder, so the early grids usually go faster than that average and the last ones slower. Three habits help:

• Bank time early. Answer the easy grids briskly, check the answer against both the row and the column, and move on.
• Never leave a blank. A wrong answer costs nothing, so when the clock is close, choose an option for every puzzle that is left.
• Say the rule before you look at the options. Six options are designed to tempt; a rule you can put in words picks one of them.

If the clock runs out before the last puzzle, the ones left unanswered simply score nothing, which is why a guess on each is worth more than a blank.

The Mensa Norway test and a supervised Mensa test, side by side

What Mensa is

Mensa International is a society founded in 1946 whose only membership criterion is a score within the upper 2% of the general population on an approved intelligence test. National groups such as American Mensa and British Mensa run their own admission sessions and social activities. The 2% line corresponds to about the 98th percentile, which on a mean-100, SD-15 scale is a score of about 130, and about 132 on an SD-16 scale.

How admission works

There are two routes. The first is a supervised admission test session run by the national organisation, with a proctor, standard timing and a fee. The second is submitting evidence of a qualifying score from a previous professionally administered test; American Mensa publishes the list of tests and scores it accepts. In both routes the score comes from a controlled administration, which is why an online quiz result is never accepted.

Why an online score is never a Mensa score

The 2% line is a percentile on a norm sample, and a percentile is only meaningful when the test, the timing, the conditions and the reference group are all controlled. An unsupervised browser test controls none of them: screen size, interruptions, a second attempt and a familiar item format all move the number. That is why national Mensa organisations accept only scores from supervised sessions or from the professionally administered tests on their published lists, and why the Norway page describes its own result as an indication.

How to prepare

Supervised Mensa tests and most other admission tests lean heavily on figure matrices and on verbal or numerical reasoning. The most useful preparation is to learn the kinds of rule that matrices use: progression of one attribute across a row, distribution of three shapes so that each appears once per row and column, two attributes varying independently, rotation by a fixed angle, and counts. Once you can describe the rule in words before looking at the options, the options stop being a guess.

Our abstract-reasoning and matrix tests use these rule types and name the one behind every question in the answer review, so you can see which type you miss. They report accuracy, not an IQ. MyTestAtlas is not affiliated with or endorsed by Mensa, and Mensa is a trademark of Mensa International.

How our own practice puzzles are written and checked, each with exactly one correct option, is on how MyTestAtlas tests are made.

Look up a term in the test and reasoning glossary.