Learn everything about
the Intelligence Quotient (IQ)

The Department of Education of France makes a wise decision. Placing children in classes of special education would require an objective certification that it suffers from learning challenges. Binet and Simon are given the task of developing the necessary first version of an intelligence test. They will publish the first version in 1905.

Binet conceived intelligence as modifiable and wanted to use his tests to identify children that needed help. He thought that with “mental orthopedics” they would improve.

Binet publishes a revised version of the original intelligence tests. Although those tests do not mention the concept of mental age specifically, the literature assigns to him around that time the first use of the concept. His goal was to create age standards or norms against which comparison could be possible.

As such mental age would be the average performance expected of a person according to its age level.

A comparison could also be made by subtracting the chronological age of the mental age to discover persons with superior or inferior capacities.

William Stern publishes his book “The Psychological Methods of Testing Intelligence” and defends that calculating the “absolute difference between mental and chronological age is not adequate, because they do not mean the same thing at different ages”. 

Stern was concerned about the fact that intelligence neither grows with age linearly nor is it distributed equally among peers depending on the age level (at a given age the differences between peers could be smaller and at another age much bigger). He therefore proposed using a mental quotient which would result from computing mental age divided by chronological age (M.Q. = M.A. / C.A.).

During a conference about intelligence testing in the USA, prof. Kuhlmann publishes an article arguing that it was necessary to better operationalize the concept of mental age and to standardize the score through the use of an index like Stern’s. He calls the index the intelligence quotient. The IQ index would allow to make comparisons between persons of the same age and also between persons of different age groups, in both cases understanding their relative position to their groups.

Lewis Terman, the author of the first versions of the Stanford Binet test, starts using in his tests the concept of Intelligence Quotient, abbreviated as “I.Q.”. The index will be the one suggested by Stern and Kuhlman with a 100 constant (M.Q. = M.A. / C.A.* 100). Example ofa child: 15 years of mental age, while 14 year-old of chronological age, that is 15/14 * 100 = IQ of 107.

New indexes and similar concepts start to appear. McCall introduces the ideal of educational quotient (EQ), in relation to the educational age. He also suggests that the index could be calculated restricted to any specific skill. For example we could compute the reading age or reading EQ (RQ). This index therefore is a comparison between the reading age and the average reading norm of his chronological age.

Another highly interesting proposal is the “accomplishment quotient” (AQ) of Franzen. This index can be calculated by dividing the actual level of achievement by the IQ. In a school example that would be dividing the child’s academic achievement quotient by the intelligence quotient from an intelligence test. The level of achievement would be the academic age of achievement divided by the chronological age.

Let’s imagine that the child is doing badly at school and has an achievement quotient of 90, while his IQ tests indicate he is very gifted with 120 IQ. The A.Q. = AchievementQuotient/ IQ = 90/120 * 100 = 75.

This measure is so interesting because it empowers to understand how children are doing versus what they are potentially capable of. Two children with the same achievement could have different A.Q. meaning that one is using more of its capacities than the other. In other words, which children are not taking advantage of their capacities, which is why some talked about the laziness factor.

Other proposals arise like the Intelligence Coefficient and the Heinis Persona Constant, which wanted to take into account the non-linear trajectory of IQ as we age. These concepts, which are not indexes, would introduce some confusion and are the reason why some people wrongly talk about the Intelligence Coefficient to talk about IQ. 

A major evolution would occur when the renowned Thurstone advocated for the use of deviation scores and percentile ranks with the normal distribution. The normal distribution is a mathematical function that shows that most cases are in the center whereas the extremes are rare. The reason to use it is that any random variable follows a normal distribution, and so does intelligence. So for any group of people we select, people will be distributed that way (most in the center, some in the extremes).

This method solved the problem of the unperfect relationship between age and intelligence, and would soon become mainstream. Learn about this method of computation later in this page.

The Wechsler Scale becomes the first relevant intelligent test to officially adopt the use of the normal curve and its deviations method to compute the IQ score. The scale author Dr. Wechsler indicated in the test manual that using deviations “relieves us from committing to any fixed average mental age for any specific age”.

The Stanford Binet tests finally incorporate the use of the deviation computations instead of mental age indexes while still keeping the 100 + 16x scale. This will give the final boost to the deviation method, which will become the standard calculation until our present times.