List price: $21.99 (that's 30% off!)
Paul Roazen is Professor of Political Science at York University in Toronto, Canada. His books include Brother Animal: The story of Freud and Tausk, Freud and his Followers, Helene Deutsch: A Psychoanalyst's life, and Encountering Freud: The Politics and Histories of Psychoanalysis
Chapter 1, entitled "What is Psychometrics" is a brief 7 page introduction and overview of the domain of psychometrics, and the kinds of tests it has produced (e.g. measuring intelligence and personality; the need for quantitative methodology to handle test scores etc.).
Chapter 2, entitled "Scientific Measurement" is where the book departs from any other book on the market (apart from Michell's (1990) graduate-level book). Instead of the usual mechanical repetition of S.S. Steven's four levels of measurement as found in many other introductory texts, what we get here is a context for measurement in its widest sense, an exposition of the crucial nature of units within scientific measurement, and then the four levels of measurement placed into a context of Steven's operational-representational theory of measurement. This chapter is very readable, and very clear, but the content will be completely novel for many students of psychology, as they are rarely introduced at such an early stage to the nature of science and measurement within science. However, although Kline shows that psychometrics does not meet the criteria for quantitative scientific measurement, he does show that the measurement it does make is worthwhile, valuable, and productive. This chapter is an excellent feeder for tutorial/seminar work. Where using it with graduate students, it should also be augmented with chapters 1 and 2 in Kline's (1998) book "The New Psychometrics".
Chapter 3, entitled "The Characteristics of Good Psychometric Tests" is a straightforward introduction to basic statistical descriptive parameters, reliability, and validity. The chapter is conceptual in focus. No formulae are presented in this chapter (just a few are given in a Glossary). Instead, great care is taken to present the issues such as measurement error in a clear and readable fashion.
Chapter 4, entitled "Test Construction" is an outline of the major procedures and methods used with psychometrics as the means by which psychological tests are constructed. The properties of desirable items are discussed, along with the major methods of item analysis and test construction. Within this chapter are basic definitions and examples of classical item analysis, criterion keying, ipsative vs normative tests, factor analysis (with a brief description confirmatory factor analysis), and item response theory.
Chapters 5, 6, and 7, entitled respectively "Intelligence and Ability Tests", "Personality Tests" and "Motivation, Mood, Attitude and Other Tests" are chapters that variously describe the kinds of tests that have been constructed for each domain. Each chapter presents the overarching definitions of the content area, the theoretical bases for the tests, and then a brief description of some of the important psychological tests that have been subsequently created.
Chapter 8, entitled "Psychometric Testing in Applied Psychology" is a description of where psychometric tests are mainly applied in professional practice, educational, occupational/ organizational, and clinical psychology. Here, Kline tries to place psychological testing into context, describing where it has proven to be essential, useful, and sometimes misleading. But, overall, the reader is given the grand picture of the impact of psychometrics on many real-world practices. This is a nice chapter - with some fine anecdotes. Kline's experience in this area shines through.
Chapter 9, entitled "The New Psychometrics and Conclusions" is the final chapter where he draws together the main features discussed in the earlier chapters, returning first to the distinction made between scientific and psychometric measurement. Kline again stresses that psychometrics is not worthless, but that it's limitations might be seen to severely restrict any future progress in the better understanding of psychological attributes. It is here that Kline introduces his concept of the "New Psychometrics" - a way forward that takes into account many of Michell's (1997) propositions, and that seems to offer a way off the plateau on which current psychometrics now seems stranded. Kline proceeds to look at the areas of intelligence and personality, trying to see where some new radical approaches to measurement might be clues to the new way forward. This will be of great interest to many readers. The last two paragraphs of the book are pregnant with meaning (to borrow a marvellous phrase from Nunnally)
So, there you have it. A very nice introduction to psychometrics, but also a logical exposition of the scientific context within which psychometrics is embedded. A unique book written by a unique professor of psychometrics. I heartily recommend it to those new to the area who wish to begin to understand psychological measurement at a level beyond mere operational knowledge.
-References-
Kline, P. (1998) The New Psychometrics: Science, Psychology, and Measurement. Routledge. ISBN: 0-415-18751-6
Michell, J. (1990) An Introduction to the Logic of Psychological Measurement. Lawrence Erlbaum. ISBN: 0-8058-0566-4
Michell, J. (1997) Quantitative science and the definition of measurement in Psychology. British Journal of Psychology, 88, 3, 355-383