How This Score Is Calculated
An honest explanation of what your number means, how we get it, and what it doesn't mean.
The Format
The test is 35 progressive matrix items, each showing a 3×3 grid with one piece missing and six candidate answers to choose from. You have 25 minutes to complete as many as you can. The items are culture-fair: no vocabulary, no general knowledge, no arithmetic — only the ability to spot the rule governing shape, rotation, size, count, and layering across each row and column. Difficulty rises steadily from the first item to the last, following the same progressive-matrix format used by well-known nonverbal reasoning tests.
How Scoring Works
Your raw score — the number of items you got right — is converted to an IQ number with a straightforward formula: IQ = 100 + 15 × (raw − 18) / 6. That places an average performance (18 correct) at IQ 100, and each additional correct answer moves you roughly 2.5 points along the scale. The result is clamped between 60 and 145. A perfect 35/35 is pinned to the 145 ceiling — the formula alone would land a perfect score around 143, but we round a flawless run up to the top of the scale. Your percentile is then derived from the standard normal distribution (mean 100, standard deviation 15), the same statistical curve used to describe IQ scores generally.
What "Calibrated" Means Here
We did not run a proctored study on thousands of test-takers to norm this scale — that's an expensive, months-long undertaking that few online tests actually do, whatever they claim. Instead, our scale is design-anchored: each item is hand-built into one of six difficulty tiers, from trivial to brutal, arranged in a deliberate ramp — every correct answer counts the same, but how far you climb that ramp before the items outpace you is what sets your raw score. The raw-to-IQ formula is calibrated against that designed structure, not against a norming study. That makes your result a strong, internally consistent estimate of your pattern-recognition ability — not a clinical measurement. Treat it the way you'd treat a well-built practice test, not a diagnosis.
The Mensa Comparison
When we tell you how close you are to a high-IQ society's admission bar, we're comparing your score to the 98th-percentile threshold — IQ 131 on a standard-deviation-15 scale — that organizations like Mensa use to qualify members. We are not affiliated with Mensa in any way; this is a factual comparison of where your number falls on the same statistical scale, not an endorsement, referral, or guarantee of acceptance into any society.
Retakes and Practice Effects
If you take the test again, expect your score to go up — this is a well-documented practice effect, not a sign you got smarter overnight. You'll recognize the pattern types faster, and pure repetition partially inflates the result. We never reveal which answer was correct on any item, on purpose: it keeps the test fair for everyone who takes it, including you the second time.
Limits
Your score on any given day depends on more than your reasoning ability — fatigue, a cramped phone screen, or a noisy room can all pull the number down. Take the result as a snapshot, not a verdict. If you need an official, clinically valid IQ score, only a trained psychologist administering a proctored test in a controlled setting can give you one.