What is Psychometrics?

What is Psychometrics?

Overview

The tagline of the Psychometric Society says that the Society is devoted to the advancement of quantitative measurement practices in psychology, education and the social sciences. This is a very general description of psychometrics, but we emphasize the word quantitative in the previous sentence. Some people take a more clinical view of psychometrics, emphasizing the administration and application of psychological scales. But scale administration is not a particular emphasis of this society.

As a way to elucidate our definitions and perspectives on the term psychometrics, a few psychometricians with different research orientations have provided definitions below. These definitions are by no means official but they rather demonstrate the individual differences in the way psychometricians think about their profession.

Henk Kelderman (Leiden University, VU Amsterdam)

Measurement and quantification is ubiquitous in modern society. In early modernity, the scientific revolution provided a firm scientific basis for physical measures like temperature, pressure, and so on. In the late nineteenth and early twentieth century, a similar revolution took place in psychology with the measurement of intelligence and personality. A crucial role was played by Psychometrics, initially defined as “The art of imposing measurement and number upon operations of the mind” (Galton, 1879, p149). Since 1936 the Psychometric Society has been at the forefront of the development of formal theories and methods to study the appropriateness and fidelity of psychological measurements. Because measurement in psychology is often done with tests and questionnaires, it is rather imprecise and subject to error. Consequently, statistics plays a major role in psychometrics. For example, members of the society have devoted much attention to the development of statistical methods for the appraisal of noisy measurements whose outcomes are considered indicators of attributes of interest that can not be directly observed.

Today, psychometrics covers virtually all statistical methods that are useful for the behavioral and social sciences including the handling of missing data, the combination of prior information with measured data, measurement obtained from special experiments, visualization of statistical outcomes, measurement that guarantees personal privacy, and so on. Psychometric models and methods now have a wide range of applicability in various disciplines such as education, industrial and organizational psychology, behavioral genetics, neuropsychology, clinical psychology, medicine, and even chemistry.

In the future we will have more personal data then ever before thanks to improved instrumentation, like brain scanning and genome sequencing, as well as the growth of the internet and computing power. Data collection now surpasses our ability to harvest and interpret its complexity. It is expected that whole economies will grow around the analysis of data, both commercial and scientific. The importance of innovations in measurement and statistics and ways to meaningfully summarize and visualize data is expected to grow along with it. The Psychometric Society is geared up to be a major player in providing these innovations.

Galton, F. (1879). Psychometric experiments. Brain: A Journal of Neurology, 11, 149-162.

Peter Molenaar (Pennsylvania State University)

Psychometrics is the approximation of latent psychological processes by means of stochastic analysis at both the individual and population levels.

Denny Borsboom (University of Amsterdam)

Psychometrics is a scientific discipline concerned with the construction of assessment tools, measurement instruments, and formalized models that may serve to connect observable phenomena (e.g., responses to items in an IQ-test) to theoretical attributes (e.g., intelligence). For example, theoretical constructs have been defined as domains of observable behaviors of which item responses form a sample (generalizability theory), as psychological attributes that act as common causes of item responses (latent variable theory), as expected values of a test score (classical test theory), as mappings of observable relations into a numerical system (scaling theory), and as systems of mutually reinforcing factors (network theory). Such models present conceptual, substantive, and statistical problems that psychometricians aim to analyze and solve. Because many of the questions that psychometricians study transcend disciplinary boundaries, and concern general issues of measurement and data-analysis, the boundaries of the discipline are fuzzy; psychometrics is especially closely intertwined with methodology and statistics. Psychometric techniques are widely used across the sciences, and have found applications in educational testing, behavior genetics, sociology, political science, and neuroscience.

David Thissen (University of North Carolina at Chapel Hill)

The use of the adjective “psychometric” in the sense of the name of the Psychometric Society dates back at least to Francis Galton’s (1879) essay in Brain entitled “Psychometric Experiments”; the opening line of that article is that “Psychometry, it is hardly necessary to say, means the art of imposing measurement and number upon operations of the mind …” Galton’s “experiments” were essentially introspective, but he treated their results “statistically” (meaning he reported counts; it was, after all 1879!). The words “psychometric” as an adjective and “psychometrics” as a noun referring to the field of study came into increasingly common use as psychology developed, reaching prominence as the name of the subdiscipline with the foundation of the Psychometric Society in 1935 and the publication of Guilford’s (1936) Psychometric Methods.

Guilford’s (1936) Psychometric Methods covered a wide variety of topics, from psychophysical methods and psychological scaling through correlation and regression to procedures for the analysis of data arising from mental tests, and factor analysis. In the preface, Guilford (1936, p. xi) wrote “The name “Psychometric Methods,” too long restricted to clinical tests and the like, is surely broad enough to encompass appropriately all the topics just mentioned.” The journal of the Psychometric Society, called Psychometrika, spelled with a k (presumably) with a nod toward Galton and Pearson’s similarly-named Biometrika), began publication with the foundation of the society; it included articles on all of the topics in Guilford’s book. For most of the twentieth century, psychometrics was defined well by the subheading that appeared under the title of Psychometrika from its inception in 1936 until a cover redesign in 1984, “a journal devoted to the development of psychology as a quantitative rational science.”

The practitioners of the science of psychometrics were known as psychometricians; however, that term was also used to refer to practitioners who administered psychological tests in educational and clinical practice. Around 1980, psychometric graduate training programs began to change their names to avoid confusion with that alternative meaning, and to be more inclusive of an ever-increasing scope; most often the nomenclature used was “quantitative psychology.” In 1984, the cover of Psychometrika was redesigned, for the first time setting aside Thurstone’s hand-ruled cover art; the time-honored title was retained, but the subheading was changed to read “a journal of quantitative psychology.” The field is now most often referred to as “quantitative psychology” but its flagship journal remains Psychometrika. Jones and Thissen (2007; see below) summarize the history of the subdiscipline in the opening chapter of a relatively recent book entitled, curiously enough, Psychometrics.

Galton, F. (1879). Psychometric experiments. Brain: A Journal of Neurology, 11, 149-162.
Guilford, J.P. (1936). Psychometric Methods. New York, NY: McGraw-Hill Book Company. 

Videos and Further Reading

Video lectures can be found here
More information on the definition, history and the future of psychometrics can be found in the following papers.

Jones, L. V., & Thissen, D. (2007). A history and overview of psychometrics. In C.R. Rao and S. Sinharay (eds). Handbook of Statistics, 26: Psychometrics (pp.1-27). Amsterdan: North Holland.
A free version of this chapter can be obtained here.

Stout, W. (2002). Psychometrics: From practice to theory and backPsychometrika, 67(4), 485-518. A free version of this paper can be obtained here.

Groenen, P. J., & Andries van der Ark, L. (2006). Visions of 70 years of psychometrics: the past, present, and futureStatistica Neerlandica, 60(2), 135-144.

Borsboom, D. (2006). The attack of the psychometriciansPsychometrika,71(3), 425-440.

 

Log in