GALTON: THE FIRST PSYCHOMETRICIAN?

Larry H. Ludlow
Boston College

Ever wonder how many brush strokes it takes to create a painting? Or how to measure boredom, attraction to the opposite sex, the efficacy of prayer, or the intelligence of earthworms? Sir Francis Galton wondered about these things and set out to develop procedures and instruments by which such questions could be answered and replicated. In fact, he counted everything that appeared to have any form of regularity.

He counted brush strokes while sitting for his own portrait at two different times in his life . He counted spikes of flowers on trees , the fidgets of persons sitting through a boring lecture, the "instances in which men who are more or less illustrious have eminent kinsfolk" , facial features , fingerprint characteristics , earthworms on a rainy sidewalk , degrees of vividness of mental imagery, instances of phantasmagoria, points of similarity among twins , variations in weather conditions , numbers of attractive, indifferent and repellent looking women , causes of snoring, and on and on. He seems to have always carried a notebook and some type of ingenious device capable of pricking a piece of paper by which he recorded, unobtrusively, various aspects of events occurring around him. He even performed arithmetic by taste and smell.

What, you might reasonably ask, is the purpose of this article? It was written because it provides some relatively obscure, yet fascinating, information on the early history of psychometrics. For some years now I have taught a course in psychometrics. An important feature of the material covered in the course is the historical context within which the models and methods we employ have evolved. However, my lectures never included anything about Galton other than his development of regression and correlation .

My approach to the history of psychometrics is fairly standard, I think. It begins with the classical German psychophysics of the 1800's with Weber, Wundt and Fechner, moves into the 1900's ability testing movement with Cattell, Binet and Spearman, and then into the psychological scaling methods associated with Thurstone. Modern test theory texts are introduced where standard presentations include something like "the field of psychometrics has a history of growth and development extending over some 75 years since the early work of Binet in France and Spearman in England" (Thorndike, 1982, p 1). And "psychometric methods" is simply defined as "procedures for psychological measurement" (Guilford, 1954, p 1). Fairly standard stuff.

But, while working on a project tracing the role that residuals have played in the evolution of scientific models, I stumbled across some early research of Galton's . In particular I became intrigued with his reference to "psychometric experiments" and I subsequently set out to track down the original use of the word "psychometrics." That effort resulted in this paper.

Galton's interests in mental operations led him to propose a "new instance of psychometry" (Galton, 1879, p 149). In his article, "Psychometric Experiments," he defined "psychometry" as the "art of imposing measurement and number upon operations of the mind". He then argued that "until the phenomena of any branch of knowledge have been subjected to measurement and numbers, it cannot assume the status of dignity of a science" . His work illustrated what he called the psychometric side of anthropology.

For his 1879 article he repeated an experiment in "mental operations" four times, under different circumstances, at intervals of about one month. The experiment consisted of recording the "thoughts arisen through direct association" with a list of 75 words . In other words, he conducted experiments in what we now call free-association . He threw his resulting thoughts into a "common statistical hotch-pot" and determined (a) the rate at which ideas were formed (50 per minute), (b) the frequency of recurrent associations (about one half), (c) the frequency within periods of his life that associations could be attributed (showing "in a measurable degree, the large effect of early education in fixing our associations"), and (d) the character of associations that occurred (verbal, sensory, "histrionic").

The significance of this article is that it is, I believe, the first published investigation in the field that we presently know as psychometrics . Granted, Galton's psychometric research differs somewhat from what we, as psychometricians, typically mean when we say we are conducting psychometric analyses but his work is compatible with our current approach to psychometrics. That is, psychometrics is the quantification of psychological phenomena.

What else does Galton have to offer? When addressing mental tests he states:

"There are many faculties that may be said to be potentially constant in adults though they are not developed, owing to want of exercise. After adequate practice, a limit of efficiency would in each case be attained and this would be the personal constant (italics added); but it is obviously impossible to guess what that constant would be from the results of a single trial. No test professes to do more than show the efficiency of the faculty at the time it was applied, and many tests do even less than this" (Galton (1885), in Pearson, Vol. II, pp. 371-2).

This quote contains the kernel of the classical true-score concept, including notions of reliability and validity. Note also that the quote appears 20 years earlier than the seminal work on measurement error by Spearman. Galton, the first psychometrician? Yes.

REFERENCES

Berg, I.A. & Pennington, L.A. (1954). (3rd ed.). An Introduction to Clinical Psychology. NY: Ronald.

Buchanan, R. (1854). Lectures in Neurological Systems of Anthropology. p 124.

Galton, F. (1869). Hereditary Genius. London: Macmillan.

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

Guilford, J.P. (1954). (2nd Ed). Psychometric Methods. NY: McGraw-Hill.

Merton, R.K. Sills, D.L. & Stigler, S.M. (1984). The Kelvin Dictum and social science: An excursion into the history of an idea. Journal of the History of the Behavioral Sciences, 20, 319-331.

Oxford English Dictionary. (1989) (2nd Ed). Vol XII. Oxford: Clarendon Press.

Pearson, K. (1914, 1924, 1930a, 1930b). The Life, Letters and Labours of Francis Galton. Vol. I, II, IIIA, IIIB. Cambridge: Cambridge University Press.

Thorndike, R.L. (1982). Applied Psychometrics. Boston: Houghton.

I gratefully acknowledge the persistent literature search efforts expended by Susan Henderson-Conlon.