As an interpretation of the results which have been obtained with mental tests, this book lies between the topics of deficiency and delinquency. It is an attempt to discover the significance of objective measurements of ability in connection with both of these fields. The pressing practical problem was to find out what positions on a scale for testing mental development were symptomatic of social deficiency. After working out a percentage method for conservatively indicating these borderlines for tested deficiency, it was then possible to reinterpret the test records of over 9000 delinquents who have been examined with some form of the well-known Binet Scale. The size of the problem of the deficient delinquent has thus been determined on a significant scientific plan. The outcome is a new basis for judging the current statements about this problem by those who have used the Binet scale. Scores of investigators by their tireless energy have provided data which may now be compared for many types of delinquents and in many parts of the country. Some sixty studies of deficient delinquents have been thus summarized from the point of view of psychological tests.

Closely related to the problem of the frequency of feeble-mindedness among delinquents is the question of the cause of delinquency. This has further been considered in the light of the most important scientific studies, especially those using the method of correlation. Among these researches stands out the fundamental investigation of the causes of criminality by Goring, a work which has received very inadequate attention in this country, although it involved ten years study of a group of 3000 convicts by the best quantitative methods. The careful study of these objective investigations should take the question of the relation of deficiency and delinquency out of the realm of opinion and theory. It may be expected to have an important influence upon the social handling of these problems. In this connection I have added a chapter of suggestions which have grown out of my year's study of the education of deficients and delinquents in European schools and institutions.

To determine the size of the problem of dealing with deficients, especially deficient delinquents, is a task of first importance. In spite of our more conservative basis for judging the results with tests, the necessity of caring for the feeble-minded remains the most vital problem connected with social welfare. The movement for more individual training in our schools, which has been gaining such headway, may also be encouraged by the evidence that maladjustment to school work is also definitely related to delinquency.

It is essential that we should have objective data for determining the borderline of tested deficiency among adults. To meet the present serious lack of knowledge on this point, new data were collected which for the first time afford the means of determining, by the use of a randomly selected group what is a conservative borderline of tested deficiency for those intellectually mature. These data include the Binet test records for all the 15-year-old children who resided in seven school districts in Minneapolis and who had not graduated from the eighth grade.

The urgency of plans for indefinitely segregating certain types of the feeble-minded, especially deficient delinquents, has placed a new emphasis on those quantitative aids to diagnosis. The difficulty of establishing feeble-mindedness before a court has been called to attention by both Supt. C. A. Rogers (173)^[1] of the Minnesota School for Feeble-Minded, and Supt. Walter E. Fernald (104) of the Massachusetts School. Both of these men recognize that psychological tests are the most hopeful way of improving this situation.

A fundamental feature of the diagnosis of deficiency is the plan here advocated for designating the borderlines on a scale on the basis of a percentage definition of tested deficiency. This involves the distinction of intellectual deficiency from certain rare volitional forms of feeble-mindedness, which the tests do not at present detect. This percentage definition seems to afford the best approach to a test diagnosis. It is apparent that the data are insufficient for finally establishing such a quantitative description of the lower limit for passable intellects on a mental scale. The plan, however, may be easily adjusted to new data, and meanwhile avoids some of the serious current misinterpretations of test results.

While the idea of a quantitative definition of the borderline of deficiency is not new, the percentage method seems to have certain fundamental advantages over either the “intelligence quotient” of Stern (188), the “intelligence coefficient” of Yerkes (226), or the description in terms of deviation, mentioned by Norsworthy (159) and Pearson (164, 166, 167). Several investigators, including Terman (57) and Yerkes (226), are utilizing the percentage method indirectly for describing the borderline of feeble-mindedness, but have inadequately distinguished it from the ratios. While ratio and deviation methods are possibly more serviceable for certain purposes, they are especially faulty near the borderline of deficiency, since they are affected by variations in the units of measurement and in the form of distribution from age to age. My paper on a percentage definition and the detailed plan for determining the borderline in the Binet scale, which was read at the meeting of the American Psychological Association in 1915, seems to have been contemporaneous with a similar suggestion by Pintner and Paterson (44). They, however, would restrict the term “feeble-mindedness” to tested deficiency, while I advocate the use of percentage borderlines on a test scale as symptomatic of one form of feeble-mindedness, much as excess of normal temperature on a clinical thermometer is symptomatic of disease.

Although no system of objective tests will ever dispense with the need for expert interpretation in diagnosing individual cases, still there are few who would doubt that it is desirable to reduce the option of expert judgment as much as we reasonably can. This is the scientific method of procedure. The borderline cases, however, which are often most troublesome in their delinquencies, are just those which will longest defy rigid rules. The diagnostician who wants to be as free as possible from external restraint will find in this border field of mental capacity a happy hunting ground. His scientific instincts should make him eager to discover when he leaves the mundane sphere and sallies forth into uncharted realms where he bears the full responsibility of his own opinion. Let me hasten to add that reasoning from objective data in the mass to the diagnosis of an individual case may lead to serious mistakes, unless one keeps alert to detect the exception from the general rule, and unless one understands the numerous sources of error entering into an examination. On the other hand the test results when properly interpreted afford the most important criteria on which to base a prognosis if they are considered in relation to the history of the case and the medical examination.

By the use of more conservative borderlines for raising the presumption of deficiency and also by designating a doubtful position on the scale, on the plan advocated herein, it is possible to make scales for testing mental capacity more serviceable both to the clinician and to the amateur tester. The latter may use the scales for his own information or may wish to discover whether an examination by an expert in mental development is desirable, without attempting to make a diagnosis himself. The scale may thus take a place in the study of child mentality analogous to the familiar Snellen chart in the testing of vision. For every teacher familiarity with a development scale may thus become as essential and desirable as the knowledge of the chart for eye testing. It should find a place in all progressive schools which do not have the services of a clinician.

The Binet system of tests was used for obtaining new data on groups of juvenile delinquents in Minneapolis and Pittsburgh. The use of this scale, around which the discussion centers, grew out of the necessity for immediate practical results for the clinic at the Minneapolis Juvenile Court which I was called upon to serve. In 1912, when that work began, there was practically nothing approaching norms with children for any other scale of tests. Even today it is plain that there is more data available for interpreting results with the Binet scale than with any other system of tests. While my experience would make me unwilling to advocate the Binet tests as an ideal method for building up a measuring scale, I still feel that it remains the most useful method at present for discovering the fundamental symptoms of intellectual deficiency. The percentage method, here advocated, as the best way available for determining the borderlines with a scale, would be quite as serviceable, however, with any other testing system. It has been my aim to contribute to the interpretation of the results of the tests as they are, not to perfecting the arrangement or details of the separate tests.^[2] It happens that one of the main objections which has been raised to the Binet scale, the inadequacy of its tests for the older ages, loses its force so far as the diagnosis of feeble-mindedness is concerned for those who accept the borderlines described in this paper.

Some diagnosticians may hesitate to use the Binet scale because of the criticisms it has received. Yerkes and Bridges state: “Indeed, we feel bound to say that the Binet scale has proved worse than useless in a very large number of cases” (226, p. 94). So far as this objection arises from the attempt to use the descriptions of the borderline of feeble-mindedness published with Binet scales, it will meet with a wide response. The difficulty is hardly less, as I shall show, with other scales. The definition of the borderline is certainly the vital point with any objective method for aiding diagnosis. Only by improving methods for determining the borderline can this weakness be attacked. The central contribution of this paper is directed, therefore, to this problem of the interpretation of the borderline, so that objective scales may be made more reliable for purposes of diagnosis.

In Part Two I have added an intensive discussion of the measurement of development and a comparison of the different objective methods for describing the borderline. This may well be omitted by those who are not interested in the technical aspects of these questions. To those who care only for accounts of individual lives, let me say that I am contributing nothing herein to that important field which has been covered in authoritative form by Dr. Healy (27) and by Dr. Goddard (112). They will find instead, I hope, the fascination of figures, a picture book in which probability curves take the place of photographs and biographies, in which general tendencies are evaluated and attention is focussed upon the problem of properly diagnosing deficiency and upon plans for the care of the feeble-minded, whether they be potential or actual delinquents.