Intelligence, the psychometric view


Following my previous post on The New Psychometrics by Paul Kline, here is a review of Intelligence, The Psychometric View, first published in 1991. I must admit I read it in two days, partly because I was on holidays and spent much more time enjoying the sun and swimming pool near Bordeaux (France, 33), but also because it is so much well written that there is no need to retrace one's steps.

Here a couple of quotes that I would like to comment on when I come back to this review. Emphasizes are mine.

(…) The relation of these two intelligence factors is highly revealing and it has been explicated in considerable detail both theoretically and empirically by Cattell (1971) under the name of investment theory, which can be briefly summarized. In this theory it is claimed that there develops a single, general relation-perceiving ability which is connected to the total associational neuronal development of the cortex. This is fluid ability, which is obviously highly heritable. Crystallised ability, on the other hand, develops as a result of investing fluid ability in particular learning experiences. Thus at an early age, say 2 or 3 years, fluid ability and crystallised ability are highly correlated. As children grow older and undergo different experiences at school and in the family, so, clearly, fluid ability and crystallised ability become less highly correlated. The bright and well-adjusted child who attends a good school and receive encouragement at home will invest most of her fluid ability in the crystallised skills of her culture. On the other hand, the equally bright child from a home where education is not valued and who attends a school of indifferent quality will not thus invest his fluid ability. His school performance may be far worse than a moderate child who invests all his ability at school. (p. 34)

This somewhat echoes what is traditionnaly observed in behavior genetics, namely that the influence of genes is higher in the early stages of development, whereas external factors may overcome their effect with environmental exposure.

It is not correct, as Spearman claimed, that human ability should be conceptualised in terms of one general factor, intelligence, together with a number of specifics. This picture is too simple. Rather we should see human ability as encapsulated by five factors, the second orders which have been described. Of these, however, two highly correlated factors are by far the most important, fluid and crystallised ability, the latter the cultural emanation of the former. Fluid ability is essentially a basic reasoning ability, necessary for problem solution on a wide variety of problems and highly heritable, being dependant on neuronal efficiency. These two intelligence factors are Spearman's g, split by more efficient factor analysis. Thus, in summary, Cattell's delineation of the structure of ability is an amplification of that of Spearman. (p. 36)

According to Cattell's work, the five (second-order) factors are:

  • Fluid intelligence
  • Crystallised intelligence
  • Visualisation
  • Retrieval capacity or general fluency
  • Cognitive speed factor


  1. Horn, J. and Cattell, R.B. (1966). Refinement and test of the theory of fluid and crystallised intelligence. Journal of Educational Psychology, 57, 253–270.
  2. Kline, P. (1979). Psychometrics and Psychology. London, Academic Press.

Articles with the same tag(s):

Multi-Group comparison in Partial Least Squares Path Models
Data Science from Scratch
Stata for health researchers
R Graphs Cookbook
Bad Data
Data science at the command-line
Reproducible research with R
Twenty canonical questions in machine learning
Do a large amount of consulting
Stata for structural equation modeling