Intelligence Is Impossible to Define or Measure?

In this essay, I will attempt to trace the development of the concept of intelligence and the various different ways of measuring it. I will discuss, starting from the early twentieth century, how intelligence first became of so much importance and of how the knowledge and understanding of the concept of intelligence has increased throughout the century. I will briefly describe the origins of the concept of intelligence and I will also mention the most recent developments in the subject such as those of multiple intelligences and artificial intelligence (AI).

Various definitions of 'intelligence' have been produced and psychologists have so far been unable to agree on a common definition. This indicates the complexity of the subject and the diverse ways of looking at it. Some of the definitions used during the twentieth century will be mentioned.

The continuous controversy as to whether intelligence is influenced mainly by hereditary or environmental causes will also be discussed.

As concepts such as 'intelligence' are always value laden, the political and ideological consequences of intelligence testing will also be briefly explored in relation to educational and racial issues.

The issue of intelligence and intelligence testing first came about within the problems of the education system. Teachers found that some children made slower progress in their studies, and explained this in terms of 'deficient capacity'. The school administrators found this explanation too simple. They believed poor academic performance could have been due to insufficient teaching. It became of some importance as to whether the cause lay on the individual or in the instruction. According to C.J. Adcock (1965, p.181) "It was this educational problem which led to the first effective tests of intelligence".

In 1904 in France A. Binet and Th. Simon were asked to create an intelligence test that looked at the problems associated with children who could not learn. Their objective was to devise a way of assessing intellectual performance. A test was constructed to assess the performance of children that were not doing so well as the rest of their peers. Binet and Simon were the first to devise such an intelligence test in order to assess a child's ability and to detect the 'defective' children. This was to provide an answer to teachers who complained about some children who were considered to be ineducable. A child's mental age was determined by their level of error - for example lower performance on tests as compared with the age-related norm.

Garlton also produced a number of tests in connection with his interest of human heredity and is regarded one of the most important pioneers in the development of mental testing. James Mckeen Caltell also was the author of several tests; he was the first to actually use the term 'mental test'. However it was the work of Binet and Simon that helped the development of mental testing get underway. Binet's approach was the asking of a variety of questions, which could be answered at different stages of development. From this it was possible to get a measure of someone's mental age. However, if one wished to compare the intelligence of children of different ages, this proved problematic. It was argued that there would have to be something more than just mental age. For instance, the older child might have a greater mental age but be less intelligent in that his/her mental age was less than average, whilst the younger one might have a lower mental age than the older child yet have greater than the average of its age group.

Therefore given the need to compare children of different ages, Stern, a German psychologist suggested the Intelligence Quotient, known as I.Q, which is the mental age divided by the chronological age, multiplied by one-hundred. If the two ages are the same then the I.Q is one hundred and the child is deemed to be of average intelligence. If, however, the mental age is higher, then the I.Q exceeds one - hundred.

However, nowadays for any good intelligence test three characteristics must be involved: 1) the test must be reliable, in other words it must consistently give similar results. This can be shown by the test re-test method. 2) It must be valid; this means that it measures what it claims to measure. Here it may be pointed out that there are three different ways of assessing validity: content validity, empirical validity and construct validity. 3) It must be standardised; this means that it must be representative of the population in question, so the individual scores can be compared against the standardised scores. A normal distribution is normally produced from the large sample of the population tested and individual subjects are located within the distribution, which is measured in standard deviations from the mean.

It must be mentioned here that IQ tests, whether psychologists liked it or not, gave rise to discriminatory practices. Children were labelled as 'idiots', 'morons', 'imbeciles' and so on according to their IQ scores. This stereotyping led to the development of an educational system that favoured the higher scoring middle class children and penalised and excluded the low scoring lower class children. IQ tests were also used to prove the lower mental abilities of other races (i.e. blacks) and therefore to justify white superiority. This however was later shown to be due to cultural bias in the construction of IQ tests. R. Davies and P. Houghton (1995, p.156) stated that " There is no doubt that different racial groups perform at different levels on IQ tests, which were designed by and for people from European and American cultures". They also argued against the term 'racial' for differences between different national groups.

Spearman (1904) suggested a mathematical approach. Spearman defined intelligence as being a 'general ability'. He had been studying and testing children, and paid particular attention to correlations of ability that occur between specific objects. He found that some performed well in the entire range of tests given, where as others performed at a below - average level. Spearman investigated many tests in this way and found that even tests that were not supposed to be intelligence tests correlated with one another and also with the tests that were supposed to be tests of intelligence. From these results according to R. Davies and P. Houghton (1995,p.152) "...he concluded that a child's performance on any test will depend upon that child's broad level of general ability, which he referred to as 'g' (general intelligence)." Spearman also thought that

...