To measure a mental age you need two things: a bank of tasks that most children of a given age can do, and a rule for turning a child's performance into a single age number. Binet supplied the first; the formula supplied the second. Here is how the whole machine worked, step by step.
What makes the method worth understanding is how much thought went into the seemingly simple move of attaching an age to a mind. It would have been easy to test school knowledge and call the highest grade a child could handle their mental level, but Binet deliberately avoided that. Knowledge can be drilled, and a child from a bookish home would look artificially advanced. He wanted tasks that leaned on reasoning and judgement, the kind of thinking that is harder to coach, so that the resulting age reflected how a child thought rather than what they had been taught. That design choice is the reason his scale worked where earlier attempts had not.
Binet's method, step by step
Collect tasks of graded difficulty
Binet and Simon assembled short reasoning, memory, vocabulary, and problem-solving tasks, arranged from very easy to hard. These were practical everyday-thinking tasks, not school knowledge, so schooling alone would not carry a child through them.
Find the age norm for each task
They tested many children and recorded the age at which most could first pass each task. A task that the average six-year-old could do was labelled a six-year-old task, and so on up the scale.
Test the child and find their ceiling
A trained tester worked through the tasks with one child, moving up in difficulty until the child could no longer reliably pass. The highest age level the child cleared became their mental age.
Compare mental age with real age
Finally the child's mental age was set against their chronological age. Ahead of their years, on par, or behind: this comparison was the whole output, and it is what the later formula made numeric.
The classic IQ formula
Binet gave the world mental age; the German psychologist William Stern gave it a ratio in 1912, and Lewis Terman at Stanford popularised the tidy multiply-by-100 form that generations learned by heart:
IQ = (mental age / chronological age) x 100
The logic is simple and rather clever. Divide how a child performs, expressed as an age, by how old they actually are. If the two match, the ratio is 1, which times 100 gives 100, the average score by design. Perform ahead of your years and you land above 100; behind, below 100. For a while, this single line of arithmetic was what IQ meant.
Worked example
Take an eight-year-old child. On the graded tasks she performs at the level of a typical ten-year-old, so her mental age is 10 and her chronological age is 8.
IQ = (10 / 8) x 100 = 1.25 x 100 = 125.
Now take a ten-year-old boy whose performance matches a typical eight-year-old. His mental age is 8, his chronological age is 10.
IQ = (8 / 10) x 100 = 0.8 x 100 = 80.
Same two ages, swapped around, and the formula turns each child's standing into a familiar-looking score. That is the entire trick of ratio IQ.
What an old-style mental-age test assessed
It helps to know what those tasks actually probed, because it was not general knowledge or reading level. The bank sampled several strands of everyday reasoning.
| Strand | What it looked at |
|---|---|
| Judgement and reasoning | Following instructions, spotting absurdities, drawing sensible conclusions from a situation. |
| Memory | Repeating digits or sentences, recalling details after a short delay. |
| Language | Naming objects, defining words, and using vocabulary appropriate to an age level. |
| Attention and problem solving | Simple puzzles, comparisons, and everyday practical problems of rising difficulty. |
The point was to sample general reasoning across a few channels rather than to test any single skill, so a child could not simply be coached to a higher age. That breadth is one reason Binet's approach outlived his own lifetime, even after the ratio scoring was abandoned.
The tasks were also arranged in a careful ladder. At each age level sat a small cluster of items that most children of that age could pass but most younger children could not, so the level a child reached was not a single lucky answer but a run of tasks cleared in a row. A tester would keep going until the child began to fail consistently, then fix the mental age at the last level solidly passed, sometimes adding partial credit for scattered successes above it. This laddered, ceiling-finding procedure is still recognisable in modern test administration, which is a quiet testament to how well the original design held up.
Why the formula breaks down for adults
The ratio works beautifully for children because their abilities climb steeply and fairly evenly year on year, so a mental age genuinely tracks against real age. In adulthood that assumption collapses. Cognitive ability does not keep rising at the same pace; on many measures it plateaus in early adulthood. Mental age, in effect, stops climbing.
The absurd consequence: if an adult's mental age sits at a plateau of, say, 30 while their chronological age keeps rising, the ratio quietly falls every birthday. At 40 the same person would score lower than at 30, and lower still at 60, purely because the denominator grew. Nothing about their thinking changed, yet the formula would declare their IQ in steady decline. That is plainly wrong, and it is exactly why the ratio was dropped.
The fix, introduced by David Wechsler, was the deviation IQ: instead of dividing one age by another, you compare a person's score with the spread of scores among people of their own age, and express where they fall. An average performer scores 100 at any age, and the awkward drifting ratio disappears. Mental age remains a useful teaching idea and still appears in some child assessments, but as a scoring engine for everyone it was retired. The research page covers that shift in more depth.
It is worth appreciating how neatly the deviation approach sidesteps the ratio's trap. Rather than treating age as a denominator to divide by, it treats age purely as the group you are compared within. Your raw performance is set against the distribution of everyone in your own age band, and your score simply reports where you sit in that distribution: right in the middle earns a hundred, comfortably above the middle earns more, and so on. Because each age band is scored against itself, there is no rising denominator to drag the number down, and the score stays meaningful whether you are twenty, forty, or seventy. The same person, tested at different ages, no longer appears to decline for purely arithmetic reasons.
Why the ratio still gets taught
Given all this, you might wonder why the mental-age formula is still explained at all. The answer is that it remains the clearest on-ramp to understanding what any intelligence score is trying to do. The ratio makes the underlying question vivid: how does this person's performance compare with what is typical for their age? Deviation IQ asks exactly the same question, just with better arithmetic underneath. Learning the ratio first, seeing precisely where it breaks, and then meeting the deviation IQ as the repair, is one of the most efficient ways to genuinely understand modern testing rather than just memorising that average is a hundred.
Where to go next
Now that you have seen the method, the natural next question is what the comparison of the two ages really tells you. That is the subject of mental age versus chronological age. Or if you would rather just play, the mental age test gives a light-hearted, non-scientific take.
Sources
- Binet A, Simon T. Methodes nouvelles pour le diagnostic du niveau intellectuel des anormaux. L'Annee Psychologique. 1905;11:191-244.
- Stern W. The Psychological Methods of Testing Intelligence. 1912 (English translation, Warwick & York, 1914).
- Terman LM. The Measurement of Intelligence. Houghton Mifflin; 1916.
- Wechsler D. The Measurement of Adult Intelligence. Williams & Wilkins; 1939.
This page is educational. Popular mental-age quizzes are for fun and self-reflection, not diagnosis or a measure of intelligence.