A HISTORY OF THE QUESTIONNAIRE METHOD OF RESEARCH IN PSYCHOLOGY. By ROBERT H. GAULT, Washington College, Md. A large proportion of current psychological literature is based upon data obtained by means of the questionnaire. By many this is regarded as a reliable, scientific method. This consideration alone is a sufficient excuse for the present paper, the purpose of which is: (1) to trace the development of the method, and if possible to find its origin, even though in the attempt we should be led outside of what is, on all sides, recognized as the science of Psychology, and into that of statistics or any other department of knowledge whatsoever; (2) in tracing this development, to point out the various fields of psychological enquiry in which the method has been or is applied. No detailed analysis of the results of the use of the method is contemplated. First of all our search for its beginning takes us into the history of statistics. The questionnaire is not necessarily a written list of questions. As long as there have been governments and tribes there have been statistics obtained by personal enquiry or by means of a written questionnaire. Perhaps the first body of numerical facts was an enumeration of fighting men. In 1835 an epoch-making work appeared in France which increased and broadened the interest in statistical knowledge throughout Europe and England. This was Quetelet's "Essai de Physique Sociale," or "Sur l'homme et le developpement de ses faculties." (38.) It was the author's aim to apply the methods founded on observation and calculation to the moral as well as to the physical sciences. This is the beginning of statistics of morals inasmuch as it first takes account of such phenomena as are under the control of man's will. It was translated into German by Riecke (1838), and into English by Knox and Smident (1842). Drobisch, a Herbartian, in 1849 attracted the interest of the scientists of the world to the new science of statistics by an article in "Gersdorf's Reportorium," Leipzig. Finally, in 1864, Adolph Wagner (46) added his powerful influence toward establishing the validity of the statistical method in the study of social phenomena, which of course include voluntary acts in endless variety. Among the earliest acts of the Statistical Society of London, which was founded in 1838 (the date of the German translation of Quetelet's work), was the appointment of committees to enquire into industrial and social conditions. (30, p. 5.) One of these committees, in 1838, used the first written questionnaire of which I have any record. The committee-men prepared and printed a list of questions "designed to elicit the complete and impartial history of strikes." Earlier than this, however, as in Bristol in 1831 (30, p. 86), house to house canvasses were made to ascertain the condition of the poor. Other committees attempted by the same method, in 1841 and 1842, to determine the state of education among the poor in Kingston-on-Hull. It is probable that there were some inconspicuous applications of this method at an earlier date than the preceding, for instance, toward the close of the 18th century when the science of statistics began to assume its modern form. Perhaps much earlier than this, for Tacitus tells us that Augustus was the statistician of his 80,000,000 subjects. The ancient Chinese produced elaborate statistical works. It would not be surprising if, in some instances, they collected their material by a method closely analogous to that which we have under discussion. But inasmuch as the questionnaire has been of conspicuous service to Psychology and Education only within the period following the beginning of Moral-Statistik with Quetelet, I have thought it would be unprofitable to make a serious attempt at going behind that event. Edwin Chadwick (10), in 1864, believed that there was a need for knowledge of large groups of children in order to enable school-men to evaluate systems of education. Accordingly he anticipated some modern investigations. He says: "In order to ascertain the results of particular methods of instruction, I sent in some parishes, circular letters to the employers of children, requesting answers to questions as to their experience of them and as to any defaults attributable to education or training for which remedies were needed." He found great differences in the results of teaching among different schools even though the subject matter of the curricula was uniform. The "conditions and manners" of the teachers varied, however, and in this was the secret of the widely differing results of instruction. He proposed a system of school visitation, and declared that from such a plan as he offered "there are already derivable educational statistics. of a new value to determine the results of educational and training power." We have no positive evidence that the English statistical work arose from continental influence. It is possible, however, that the translations of Quetelet's work, mentioned above, stimulated investigation in Britain. Neither is there any evi V dence up to the present that the new science of statistics is supplying a method for Psychology, unless we describe Chadwick's problem as a psychological one, and can show that he was led to it under the influence of statistical science. There is no proof of this further than that the results of his research appeared at a time when the method of statistics was young and flourishing and was attracting wide attention. As far as its time of appearance is indicative of anything, however, it might have been suggested by Berthold Sigismund's "Kind v und Welt." The place of this latter work in the development of the questionnaire method will be considered presently. The method we are discussing was very early applied in the study of childhood. It was Helvetius' (26) (1715-1771) belief that the genius and the dunce are alike in infancy. The influence of environment in either case produces the final product. If so the genesis of both types is open to observation. Carlyle's suggestion, made a generation later, is, therefore, to the point: "Of a truth it is the duty of all men, especially of all philosophers, to note down with accuracy the characteristic circumstances of their education, what furthered, what hindered, what in any way modified it." (9.) In 1787 Tiedeman first obeyed the spirit of such a suggestion as this when by systematic observation he gleaned material for a little volume which has been translated into English under the title, "Record of Infant Life" (45). That a strong desire for life histories was abroad at the time is shown in the "Allgemeinen deutchen Bibliothek" (1) (1765-1805), which expressed a wish for a history of all that which has occurred in the soul of a child from its first sensation, or rather from its first movement, to the first use which it makes of its reason. The fulfillment of this wish, it was declared, would spread a brighter light over the ways of the soul than all systems which the philosophers have constructed from the beginning of the world. All this, together with the interest in children which Pestalozzi had aroused, was sufficient to prepare the way for the work of Berthold Sigismund (1819-1864). Sigismund was a man of high scholarly attainments. A physician, his health compelled him to give up the practice of his profession. He became a teacher and was always strongly inclined toward the natural sciences, though he made wide excursions into the languages, music, and other forms of art. As a teacher he was interested in children and devoted much of his time to a careful, first-hand observation of their language and movements. Striking individual variations, which appeared to him in the course of these observations, showed the desirability of wider studies. It is in obedience to a scientific motive that he writes in his introduction to "Kind und Welt"; "I concluded therefore to put together the results of my observations and to send them in copy to several mothers of good judgment, in order to obtain through them a collection of methodical biographies of children, from which, by induction, I might derive those laws of human development for which I had sought in vain in books." (41, p. 10.) It was the especial purpose of that little book, as Sigismund says further in the introduction, to create an interest in the special observation and study of childhood. His motive, I have said, was scientific. The method he used and proposed for extended applica-✓ tion he regarded as statistical. He says: "Concerning the causes of the early or late unfolding of the mind in particular children, as yet we know nothing at all, naturally, since as yet no one has taken the pains to do the preliminary statistical work which is so easily executed. Statistics has in our time done extraordinary things. It has shown how many pounds of flesh a man consumes on the average in England or in Prussia; how long he lives here or there, on the average; how many crimes are committed here or there. If it would also once turn its searching eye upon the development of man!" (41, p.39.) It was Sigismund's wish that a scientific society should be established to apply statistical science to this purpose: the investigation of man's development. The publication of "Kind und Welt" marks an important ✓ epoch in the development of the questionnaire method. In it there is suggested a definite plan for other students to follow. It is not to be inferred, however, that all who employed the method within a few years following, did so under the influence of Berthold Sigismund. A few years later G. T. Fechner used what seems to be the first questionnaire syllabus applied in the study of adult human psychology. After having observed striking personal differences in memory images and imagination, he determined to conduct a statistical investigation of these images and their individual variations. Enquiries were made (in exactly what form he does not relate) of several well known German scholars, whose names, with an abstract of their replies, are given in the Elemente der Psychophysik. (14.) This was in 1860. Just seven years later (three years after Chadwick's questionnaire study of educational conditions) Charles Darwin undertook the investigation of a very definite problem by the same method; "Expression of the Emotions in Man and Animals." (12.) The problem was to trace out the genesis of the expression of emotions: is it innate or acquired? It was practically impossible for the author to obtain the data 1 he needed for the answer of his question by direct observation. He adopted the only alternative that presented itself. Others whom he regarded as reliable observers, at home and abroad, among both civilized and barbarous peoples, made the observations for him, following the directions that were given in his famous questionnaire. The questions were direct. The respondents were required not to trust their memory but to report only what they observed at the time of answering each question. This requirement, the author says, was fully met by his correspondents. Thus they avoided the serious error which inevitably attaches to reports from memory. One cannot determine from Darwin's discussion of his method in the introduction to his book whether he owes the plan to suggestions from an outside source or whether it is his own device. It is certain, however, that he began his study in 1838, more than thirty years before the publication of his results, and at a time when the statistical method was coming into prominence among scientific men. Francis Galton applied the method in a new field of Psychology. Two standard works from his hand, and at least one of minor importance, are based upon questionnaire returns. 1874 His "English Men of Science" (15), Galton himself tells us was undertaken after he had read M. de Candolle's History of Science and Scientists. (13.) This is a statistical enquiry into the life histories of about two hundred French scientists of the two centuries preceding its publication. The data on which it was based were collected entirely from biographies and society records. It happened that, at the time of its appearance, Galton was working more extensively along a parallel line with the purpose of supplementing his earlier work, "Hereditary Genius." Because of de Candolle's dissent from some of Galton's views on heredity, the little volume, "English Men of Science," was put into print, based upon entirely "new and trustworthy materials." He had addressed copies of a long and tedious questionnaire to scientific men in England for the purpose of obtaining an account of their life histories. The majority of his questions were necessarily indirect. He does not discuss them extensively, neither does he speak of the different senses in which they may have been comprehended by his respondents. Galton did not anticipate that his method would prove itself so delicate as to show what chances there are that the descendants of a famous physicist will become eminent in the science of physics. He would be satisfied should it show in a more general way that those descendants were likely to become eminent in any field whatever. In this he has apparently succeeded. Th. Ribot, at any rate, says (40) that, in this work, Galton |