Ian J. Deary's new book is about the biology of individual differences in intelligence. He views psychometrics - the measurement and statistical study of psychological traits, such as intelligence - as being between social variables (such as school attainment, income, accident rates, etc.) and the biology of the brain, with each field building on a base of knowledge from the lower levels. From psychometrics one can look upwards at applications of psychological tests, or one can look downwards onto the biological factors that affect intelligence. Looking down is what this book does.
The basic approach is historical, with most modern research described in the last chapters. I would suggest reading the first chapter and then jumping to the later chapters that describe the more modern research. Some of the early writings are mainly of historical interest. These merely serve to demonstrate that the same questions and hypothesizes have been of interest for centuries.
The introductory chapter deals with a number of topics that will be familiar to the specialist in intelligence. One is the existence of a general factor. When mental tests are given to the same subjects, the subjects' performances on the different tests are positively correlated. The person who does well on one test generally does well on other tests, and the poor performer on one test is likely to be a poor performer on other tests. This is not what would be expected if solving different types of problems were a learned skill with those better at solving a particular type of problem being better because they had had more practice. Since time spent practicing solving one type of problem is time not spent at practicing on other types of problems, one who was well practiced at one type of problem would tend to be less practiced at another type of problem. In this case, there would be an inverse correlation between performances on different type of problems. However, a positive correlation is observed. This simple observation suggests that individual differences in practice or in study time do not explain individual differences in performance.
A brief discussion is given of the many social variables intelligence helps predict such as school performance, income, social status etc. Particularly striking is the data from Hunter and Hunter (1984) on the validity of common hiring criteria. Best at predicting job performance were mental ability composites (.53). This was even better than job tryouts and far exceeded such measures as education (.10), academic achievement (.11), or interviews (.14). Naturally, some of the evidence for the importance of intelligence comes from The Bell Curve (Herrnstein & Murray 1994), which documents the large number of socially important variables that are influenced by intelligence. (A summary of the findings are available in Miller 1995)
The introductory chapter also summarizes the evidence on the heritability of intelligence. The traditional methods of behavioral genetics using twins and adoptees show that there is substantial genetic influence on intelligence, which appear to account for most of the variability between people in intelligence.
After the introductory discussion, genes are not discussed again until the next to last chapter, which deals with the latest developments in biological research. This chapter (written with coauthors Carl and MacLullich) discusses the quantitative trait loci project. This is an effort to identify specific genes that affect intelligence (or at least are markers for these genes). The basic methodology here is to look for genetic markers that are shared by individuals of high intelligence. To minimize the number of subjects studied the latest research has focused on very high intelligence (gifted with an IQ averaging 136) children, which are compared with normal children. So far evidence has been found that a variation in the insulin-like growth factor-2 receptor (IGF2R) gene is more common in the high IQ groups than in the normal IQ groups and that the effect is unlikely to be due merely to chance. …