An aptitude test measures something real. The problem is not there: it is in the gap between what it measures and what people believe it measures.
Creativity
Producing unexpected and useful ideas draws on different mechanisms from solving problems with one right answer. It is structural: an aptitude test has, by construction, one correct answer per question. It therefore cannot assess the ability to invent a new one.
The research suggests a partial relation — some level of ability appears necessary for creative output — but beyond that the correlation weakens sharply. What you do with a space of possibilities is not the same as your ability to traverse a constrained one.
Practical judgement
Deciding whether to change job, defusing a conflict, judging whether someone is reliable, picking the moment to say something. None of that appears in a mental rotation item.
This is probably the widest gap between the test and real life, and the one everyone has observed: people who are brilliant at formal tasks and consistently wrong-footed in human situations, and the reverse.
Persistence and conscientiousness
These are personality traits, not aptitudes. They appear in no IQ test — and they predict educational and occupational outcomes at least as well as cognitive ability, sometimes better.
A moderately quick student who works steadily overtakes, across a whole education, a sharp one who never starts. That is not a consoling moral: it is what the data show.
Long-term memory and knowledge
A well-built test deliberately minimises dependence on these — that is the whole point of matrices and series, intelligible without specialist vocabulary.
What is called crystallised intelligence, the structured body of what you know, is largely out of scope. Which is paradoxical, because it is often what the people around you mean by “intelligent”.
Sustained attention
A test runs twenty minutes to two hours. Sustaining intellectual effort over six months is a different matter, governed by other mechanisms — motivation, emotional regulation, organisation.
“Emotional intelligence”
Care is needed here, because the term covers very different things depending on who uses it. In its scientific form — measured by performance tasks on recognising and regulating emotion — it is a studied construct, modestly correlated with verbal ability and distinct from the rest.
In its popular form, the one of self-report questionnaires and corporate training, it is often a repackaging of personality traits. The promises attached to it run well beyond what the evidence supports. It remains true that an IQ test measures none of it.
About the correlations people cite
You regularly read that IQ predicts occupational success or income. That is accurate, but it is worth understanding what it means.
The established correlations are real and moderate. A moderate correlation describes a tendency at group scale: compare ten thousand people and those with higher scores will on average do better on those criteria.
But the same correlation becomes a poor predictor the moment you apply it to one person. The vast majority of individual variation is explained by something else — social background, the luck of who you meet, health, choices, effort, era.
Confusing those two levels of reading is the commonest error on this subject, and it is made as often by people who overrate tests as by those who dismiss them entirely.
So what is being measured?
Something narrow and real: the ability to extract an abstract rule, transpose it, manipulate relations between symbols, and hold several constraints at once, under time pressure.
That is useful. It corresponds to part of what people mean by intelligence in ordinary life. But only part, and a twenty-minute test gives only an estimate of it.
Which is why our test shows a range rather than a number, and why its name — Just an IQ Test — is as much a warning as a title.