Free Classical Intelligence Test

The classical four-subtest intelligence test, explained

Almost every serious intelligence battery of the last century has been built the same way: several short subtests, each aimed at a different kind of reasoning, added together into one general score. The reason is not tradition but arithmetic. Any single kind of question measures general ability plus a large helping of one narrow skill, and the narrow part is noise if what you want is the general part. Add four different question families together and the narrow parts partly cancel, while the general part accumulates. That is why a mixed battery of thirty items is worth far more than thirty items of the same type, and it is the whole design principle behind this page.

What each of the four sections asks of you

The verbal section is about relations between meanings, not vocabulary size. You are asked to complete analogies, to pick the word that breaks a group, to choose the word that reverses a meaning, and to finish a sentence whose connective quietly tells you which direction the missing word must point. The words themselves are ordinary; the work is in seeing precisely how two of them relate.

The numerical section mixes sequences with short quantitative problems. Half the items give you a run of numbers built from a hidden rule — a doubling, an alternating step, a growing difference — and ask what comes next. The other half are the sort of proportion, percentage and average problems that reward a clear head rather than fast fingers. Nothing needs a calculator, and everything can be done in two or three lines.

The spatial section is made of three-by-three matrices. Each grid follows a rule about shape, count, shading or rotation, and one of six tiles completes it. These items carry no language at all, which makes them the fairest part of the test if English is not your first language, and the part that most closely resembles what psychologists call fluid reasoning.

The logical section asks whether a conclusion actually follows. Some items are syllogisms where the tempting answer overreaches what the premises support; others are conditionals, orderings, negations, or small constraint puzzles where exactly one arrangement survives. This is the section where careless readers lose the most points, because the wrong options are usually true-sounding rather than obviously false.

How the questions were built

Every verbal, numerical and logical item on this page was written from scratch for it, and the matrix items are generated from explicit rules rather than drawn by eye. In each case there is exactly one defensible answer: the distractors are near misses — the plausible over-general conclusion, the sequence value you get from the wrong rule, the tile that changes one attribute — so eliminating options is real work rather than a formality. The position of the correct answer moves around from item to item, so there is no lazy pattern to exploit.

Reading your overall score and your profile

Your raw total is mapped onto the standard scale where 100 is the population average and 15 points is one standard deviation. With thirty items the estimate is reasonably stable, but it is still an estimate: read it as the centre of a band roughly eight points wide.

The section breakdown is the more interesting half of the result. A flat profile — all four sections within one item of each other — means your abilities are evenly developed and the overall number describes you well. A profile with one section standing clear of the rest tells you something the total hides. A strong spatial score with a weaker verbal one is common among people who learned English later or who work in visual trades; the reverse is common among heavy readers. Because each section holds only seven or eight items, ignore gaps of a single question and pay attention only to gaps of two or more.

Getting a number you can actually trust

• Take it once, rested, on a real screen. The matrix items are small on a phone and squinting costs seconds you cannot spare.
• Answer everything. Nothing is deducted for a wrong answer, so a blank is strictly worse than a guess you have narrowed to two options.
• Budget your time. Fifty seconds an item is the average; spend ninety on a hard one only if you have banked the time somewhere else.
• Read the logical items twice. Most losses in that section come from answering the question you expected rather than the one printed.
• Do not look anything up and do not use a calculator. Both inflate the score and destroy the only thing it was going to tell you.

How this differs from the other tests here

The quick IQ test is a five-minute sample that places you in a broad band and nothing more. The culture fair IQ test spends all of its items on visual patterns, which makes it the most language-neutral option and the best choice if you want a single steady number. This classical test is the only one of the three that reports a profile as well as a level, which makes it the right starting point if your question is "what am I good at" rather than "what is my number". If you plan to take more than one, leave a day between them; back-to-back testing measures your stamina.

What an online battery cannot do

A supervised assessment differs from this page in ways no website can close. A psychologist controls the room, watches how you approach a problem rather than only whether you solved it, administers subtests that require spoken interaction, and compares you against a properly stratified norm sample instead of against whoever happened to visit a website. Screening tests like this one are useful for curiosity, for practice before an employer's aptitude round, and for noticing a large gap between two of your own sections. They are not a basis for a diagnosis, a school placement or a hiring decision, and anyone who tells you otherwise is selling something.