Modern Psychometrics

John Rust, Susan Golombok

• Publisher: Routledge
• Year: 2018
• eBook ISBN: 9781317724728
• Edition: 2
• Topic: Psychology
• Subtopic: History & Theory in Psychology
• Index: Psychology

The edition of this classic text has been completely revised and updated, taking into account recent developments in the field of psychometrics. Part 1 of Modern Psychometrics outlines the background, history and controversies surrounding psychological testing. Part 2 provides a practical guide for developing a psychometric test. Modern Psychometrics forms the ideal companion for those studying for the British Psychological Society's Certificates of Competence in Testing

Part one

Chapter 1

The development of psychometrics

Definitions and origins

Psychometrics today

The history of psychometrics

Galton and the origins of psychometrics

What is intelligence?

Intelligence testing and education

IQ tests and racism

Issues in intelligence testing

Twin studies of intelligence

Societal differences in intelligence

Test bias and test validity

Intelligence and moral worth

The neuropsychological view of intelligence

Intelligence and cognitive science

Alternative models of intelligence

The ethics of IQ testing

The limitations of IQ tests

Definitions and origins

Psychometrics is defined in Chambers Twentieth Century Dictionary as the ‘branch of psychology dealing with measurable factors’, but also as the ‘occult power of defining the properties of things by mere contact’. While it is the first of these definitions that we shall be dealing with in this book, there have been times in recent years when the second might have seemed more accurate as a description of current practice, particularly in debates about intelligence. It is impossible to consider the development of modern-day psychometrics without looking at the substantial influence of the intelligence testing movement in the late nineteenth and early twentieth centuries. However, the origins of the subject go back long before then.

Employers have assessed prospective workers since the beginnings of civilization, and in all probability have had consistent and replicable techniques for doing this. The earliest recorded examples of examinations for this purpose are from China at the time of the Chan dynasty around about 1000 years BC. Records show that the officials of the emperor were examined every third year within examination halls specially designed and built for the purpose. There were job sample tests that required the demonstration of proficiency in arithmetic, archery, horsemanship, music, writing, and skills in the performance of rituals and ceremonies. Formal procedures required, then as now, that candidates’ names should be concealed, independent assessments by two or more assessors should be made, and that the conditions of examination should be standardized. The pattern set down then – of a ‘syllabus’ of material which should be learned, and an ‘examination’ to test the attainment of this knowledge – has not changed in framework for 3,000 years and was in extensive use in Europe, Asia and Africa even before the industrial revolution.

Psychometrics today

Today, we experience assessment in a wide field of our activities, not just the psychological. We are tested at school to monitor our performance. We are tested at the end of school to provide us with our first academic credentials, a process that will continue throughout our learning lives. We have to pass both practical and written tests to obtain our driving licences and to be able to practise our professions. We are tested in order to gain special provision (e.g. for learning difficulties) or to obtain prizes. When we buy on credit or apply for a mortgage we have to fill in forms that are scored in much the same manner. We are tested at work, when we apply for promotion, and when we seek another job. The forms of assessment can also take on many forms – interview, examination, multiple-choice, diagnostic, practical, continuous assessment and so on. But in spite of the wide variety of applications and manifestations, all assessments share a common set of fundamental characteristics – they should be reliable, valid, standardized and free from bias. There are good assessments, and bad assessments, and there is a science of how to maximize the quality of assessment. That science is psychometrics. There is no other aspect of the field of psychology that has such an impact on individuals in their daily lives.

The history of psychometrics

Rapid scientific and social progress in Europe during the nineteenth century led to the development of several assessment techniques, most notably in the medical diagnosis of the mentally ill. However, the most dramatic impact was to come from a branch of pure science – biology. Darwin was the giant figure of the age, and his theory of evolution had considerable implications for the human sciences. In particular, his argument for the evolution of man by natural selection opened the door to human genetics based on evolutionary and biological precepts. In The Descent of Man, first published in 1871, Darwin argued that the intellectual and moral senses have been gradually perfected through natural selection, stating as evidence that ‘at the present day, civilized nations are everywhere supplanting barbarous nations’.

Darwin’s ideas of natural selection involved the intervening stages of ‘the savage’ and ‘the lower races’ at an inferior level of evolution to ‘the civilized nations’. These ideas were, however, not introduced by Darwin but rather were his natural interpretation of prevailing opinion in England in the nineteenth century. They provided justification for colonialism and the class system, and served to maintain the British Empire.

The evolution of the human intellect was of particular interest to Sir Francis Galton, who in 1869 published Hereditary Genius: An Inquiry into its Laws and Consequences. Galton carried out a study of the genealogy of the famous scientific families of the time, and argued that genius, genetic in origin, was to be found in these families (which included his own). Thus we had at the end of the nineteenth century a popular scientific view, in accord with the philosophy and politics of England at that time, that evolutionary theory could be applied to man, and that the white, English, middle-class men of letters were at the peak of the human evolutionary tree. The hierarchical theory gave inferior genetic status to apes, ‘savages’, the races of the colonies, the Irish, and the English working class, and served as a justification for the social position of the dominant group.

Galton and the origins of psychometrics

Galton is generally credited with being the founder of psychometrics. He established an anthropometric laboratory at the South Kensington Exhibition in 1883, where persons attending the exhibition could have their faculties tested for threepence, and the data generated from this and other studies provided the raw material for the development of the tools of the trade. He also developed the twin study as a technique for looking at heredity, and together with his colleague, Karl Pearson, developed the Pearson Product-Moment Correlation Coefficient for analysing these data.

In fact, the attempts to measure intellect by these early tests were a failure, as very few of Galton’s measures – visual, auditory and weight discrimination, threshold levels and other psychophysical variables – were particularly related to each other. Cattell in 1890 and Gilbert in 1894 both carried out large-scale correlational studies using university students as respondents, examining the relationship between academic grades, psycho-sensory tests and anthropometric measures such as size of brain and shape of head. While grades correlated highly with each other they showed no meaningful relationship with the physical or sensory measures (Wissler 1901). However, the techniques and models of analysis still form the basis of present-day psychometrics. Galton also explored the idea of using the normal curve as a model for the distribution of test scores.

Pearson continued to develop the mathematics of correlation, adding partial and multiple correlation coefficients and the chi-square test to the repertoire of available techniques. Charles Spearman (1904), a former army officer turned psychologist, further developed procedures for the analysis of more complex correlation matrices and laid down the foundations of factor analysis. Thus by the first decade of the twentieth century the fundamentals of test theory were in place, and used almost entirely in the development of what had come to be called ‘intelligence tests’.

What is intelligence?

The earliest pioneers in the area were generally unclear about what they meant by the concept of intelligence, and the question ‘What is intelligence?’ is still with us today. Galton (1869) believed that the key element was sensory discrimination but effectively defined intelligence as that faculty which the genius has and the idiot has not.

The discriminative facility of idiots is curiously low; they hardly distinguish between heat and cold, and their sense of pain is so obtuse that some of the more idiotic seem hardly to know what it is. In their dull lives such pain as can be excited in them may literally be accepted with a welcome surprise.

Herbert Spencer considered it to be ‘the mental adjustment of internal relations to external relations’. Charles Spearman emphasized school achievement in subjects such as Greek. It seems clear that these definitions have not arisen out of a scientific psychology but are extensions of the folk psychology of, if not the common man, the common schoolteacher. This psychology recognizes an important distinction between the educated person and the intelligent person. The former is someone who has benefited from a sound education. The latter is someone whose disposition is such that, were they to receive such an education, they would perform very well indeed. Whether a person receives such an education or not is very much a matter of social circumstance, so that a particular educated person is not necessarily intelligent, nor a particular intelligent person educated. Rather, the intelligent person was someone who could make the most of their education, and this was seen as part of the person’s ‘disposition’.

This view of intelligence was very familiar to scientists in the nineteenth century almost all would have studied Latin and Greek at school and university, including the works of Aristotle and Plato. Thus in Plato’s Republic Book V, Socrates asks Glaucon

When you spoke of a nature gifted or not gifted in any respect, did you mean to say that one man will acquire a thing hastily, another with difficulty; a little learning will lead the one to discover a great deal; whereas the other, after much study and application, no sooner learns than he forgets; or again, did you mean that the one has a body that is a good servant of his mind, while the body of the other is a hindrance to him? – Would not these be the sort of differences which distinguish the man gifted by nature from the one who is ungifted?

Thus, intelligence was not education but educability. It was perceived as being part of a person’s make-up, rather than socially determined, and by implication their genetic make-up. Intelligence when defined in this way is necessarily genetic in origin. Further underpinning for this approach came from psychiatry, where elementary tests were being developed to distinguish the insane from the imbecile, and as some of the various forms of mental defect were found to be due to genetic anomaly, so evidence was piled on presupposition.

Intelligence testing and education

Much of the early work on the measurement of the intellect was theoretical; however, applications were obvious and needs were pressing. In any society where opportunities for work or educational facilities are less than the demand, some form of selection is inevitable. If the job or educational programme is demanding in terms of the amount the applicant will need to learn before competency is reached, then there is an inclination to accept those who are seen as easier to teach, a task that could be simplified by testing for educability, or intelligence.

Alfred Binet was the first to provide an intelligence test specifically for educational selection. The main impetus came when the Minister of Public Instruction in Paris in 1904 appointed a committee to find a method that could separate mentally retarded from normal children in schools. It was urged ‘that children who failed to respond to normal schooling be examined before dismissal and, if considered educable, be assigned to special classes’. Drawing from item types already developed, Binet put together a set of thirty scales which were standard, which were easy and quick to administer, and which effectively discriminated between children who were seen by teachers to be bright and children who were seen as dull, as well as between mentally retarded children in an institution and children in ordinary schools.

Following Galton and Cattell, psychophysical and sensory tests were known to be poorly related to educability, so Binet emphasized in his tests what he called the higher mental processes: the execution of simple commands, co-ordination, recognition, verbal knowledge, definitions, picture recognition, suggestibility and the completion of sentences. He believed that good judgement was the key to intelligence.

judgement, otherwise called good sense, practical sense, initiative, the faculty of adapting one’s self to circumstances. To judge well, to comprehend well, to reason well, these are the essential activities of intelligence. A person may be a moron or an imbecile if he is lacking in judgement; but with good judgement he can never be either. Indeed the rest of the intellectual faculties seem of little importance in comparison with judgement.

The first scale was published in 1905, but an improved version came out in 1908 in which the tests were sorted into age levels, and in 1911 other improvements were made. Tests that might measure academic knowledge rather than intelligence – reading, writing, or tests of knowledge that had been incidentally acquired – were eliminated. The Binet tests and their derivatives (the Stanford–Binet in the USA and the Burt tests in the United Kingdom) were widely used throughout the world for the next sixty years for diagnosing mental retardation in children.

IQ tests and racism

Modern books on testing often contain an ambiguity of purpose concerning the use of intelligence tests with children. On the one hand they seem to be required in order to identify brighter children who may be allowed to explore their potential unencumbered by the presence of slower learners within the learning environment. On the other hand, the intelligence tests enable us to identify children with learning difficulties in order that special resources can be made available to them. Both of these issues are particularly important to debates over streaming and separate schooling. There is an apparent confusion, not without political overtones, between the idea that the bright children should not be hel...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• List of Figures and Tables
• Preface
• Part one
• Part two
• Bibliography
• Index