The Binet scale in its various forms provides only part of the objective evidence as to the mental inferiority of delinquents, although it affords the best means at present of interpreting the borderline of deficiency. Among the other investigations in which psychological tests have been tried with delinquents in comparison with normal subjects, the recent study of the Mentality of the Criminal Women by Weidensall is the most important so far as estimating the frequency of deficiency is concerned (60). It affords an admirable check upon our conclusions from the Binet examinations, since she gives in detail the results with a random group of 88 women inmates of the Bedford (N. Y.) Reformatory, which is quite comparable with the group of 200 which she tested with the Binet scale, and which we have already considered.

For our purpose, the most important comparisons are those between the group of women in the reformatory and the group of 15-year-old Cincinnati working girls tested by Woolley with the same tests. Weidensall's Table 92 shows for three tests the percentages of the Bedford women who tested below the lowest 1% of these girls. For the opposites test, 20% were below this borderline; for a test on the completion of sentences, 12%; for the memory span for digits, 29%. She also shows that 17% of the delinquent group were poorer than any of the working girls and 30.7% as poor as the poorest 5.7% of these working girls, when their mentality is measured by the number of the tests in which their ability is at or above that of the median working girl of fifteen. This 30.7% is probably most nearly comparable in ability with the lowest 0.5% of the general population.

Kelley's monograph on Mental Aspects of Delinquency, to which reference was made in the last chapter, gives the results with boys in the Texas Juvenile Training School for the completion test and his own construction test, as well as for a number of physical measurements, sensory and motor tests. He has used various data from which to provide norms for comparison. In connection with the Psychopathic Institute at the Chicago Juvenile Court, Healy and Fernald (125) have published an elaborate series of tests with suggestions as to how they may be employed for analyzing a child's mental ability and estimating his mental capacity. Schmidt has partially standardized these tests (178). Guy G. Fernald (15) tried out a dozen different tests and recommends seven of them for testing delinquents who are of adolescent age or older. Haines has sought the diagnostic value with girl delinquents of a dozen tests including Fernald's test of moral judgment. Weidensall (218), Smedley (51), Rowland (49), Porteus (45), and Whipple and Fraser (220, p. 663), have published results with certain tests tried with delinquents. With none of these tests can we adequately define the borderline of feeble-minded intellects.

There is no series of tests which has been employed outside the field of delinquency which diagnoses the borderline cases objectively so well as the Binet scale. The tests of Weyandt (219), Rossolimo (175), Rybakow (176), and Knox (134) are without definable limits based on unselected groups. Those employed by Dr. Norsworthy, while scientifically better scored for describing the borderline, were not arranged with this in view (160). Carpenter has published norms obtained with Squire's tests on 50 pupils of each age from 7 to 14. Single tests like the form board (87), Knox's cube test (134), the substitution test (1), and the A test (160) have been tried with delinquent or feeble-minded groups as well as with normal people. Under the direction of the New York Board of Charities an excellent beginning has been made in determining norms for eleven different tests (158). Stenquist, Thorndike and Trabue (54) have furnished developmental norms for several tests. Gilbert (108) and Smedley (51) at an earlier date provided age norms and deviations for certain tests. Mrs. Woolley has provided the percentile distribution for a series of mental and physical tests with 14-and 15-year-old children leaving the public schools to go to work (222) (223). In England a goodly number of different tests have been tried out on small groups or on children of particular ages (84) (63) (224). Pyle has obtained norms and variations with a series of group tests. It approaches nearest to the Binet as a developmental scale for the immature, but these tests have not been tried as individual tests and so could hardly be used safely for individual diagnosis. A graphic summary of the developmental curves for most of these tests on children will be found in Chapter XIII.

In no case do we find any tests except the Binet scales which have reached a stage of practical usefulness for the diagnosis of deficiency except as supplementary aids for checking the Binet indication with children of particular ages. The emphasis has almost universally been placed on determining the central tendencies of children of different ages and not on the lower limits of the distributions. Considering mental tests apart from the Binet scale, in all the extended literature which has been brought together in books like Whipple's Manual of Mental Tests (220), one may seek in vain for tests which have reached the position of defining the limits of serious mental deficiency. This indicates, of course, the difficulty as well as the newness of the problem, although the quantity of work that is being done shows the great interest aroused. From all of this mass of research on mental tests one may gather much that is useful in analyzing the character of a mental defect. Many of the tests admirably aid in elaborating the subjective impression of the examiner. The failure to do this systematically has been one of the main criticisms raised against the Binet scale. This and the incorrectness of the borderline described in the published scale seem to be the main objections made by Miss Schmidt to the Binet Method. She voiced the objection of the Juvenile Psychopathic Institute in Chicago to the tests as follows: “It has been the experience of the writer, and it may be added of all others who have worked in this laboratory, where practical results are demanded, that the Binet tests cannot furnish an adequate means through which to come to conclusions for the disposition, classification, or treatment of the cases which come for diagnosis” (179).

Dr. Merrill of the Seattle court also seems unfriendly to the Binet scale when he says: “Any system of tests by which alone^[30] it is attempted to classify the child as being of a given mental age involves the fallacy of pseudo-exactness, and needs carefully to be avoided” (148). Nobody would seriously urge that real exactness of definition leads to confusion. It is just the looseness of definition of borderline with the Binet Scale which has led to most of the mistakes with it. Perhaps Dr. Merrill has not discovered that the scale works just as well when used as a graded series of tests without the designation of mental ages at all. The latter is merely a convenience. On the other hand, we should agree when he says, that “no scale of tests can give a valid measure of the child's intelligence unless supplemented by a consideration of his history,” especially if he includes in the child's history a medical diagnosis.

The objection that the Binet tests do not analyze the source of the child's mental defect is of course important if one were considering whether a better scale might not be devised. It is rather beside the point, however, when one remembers that it is not the purpose of this scale to determine the causes of deficiency, but only to say whether a deficiency in general intelligence is present and to what degree. The causes of the disturbance must then be determined by an expert. Moreover, if one classifies the Binet tests as Meumann has done one may often get valuable clues as to whether the deficiency is mainly in information or in mental process. In seeking the causes of the disturbance, the expert should not overlook the standardization of the Rosanoff and Kent Association Test which has been available for delinquent, feeble-minded and normal children (174). It is one of the most important supplementary means for mental analysis which has yet been standardized for practical use. The most complete tables on children's reactions for this test have been published in a Psychological Monograph by Woodrow and Lowell.

The importance of more accurate psychological tests in studying mental disturbance is well illustrated by comparing the results that may be obtained with the Binet tests with the desultory, unstandardized tests such as one finds in Dr. Schaefer's Allgemeine gerichtliche Psychiatrie für Juristen, Mediziner, and Pädagogen (177), or Dr. Cimbal's Taschenbuch (91) prepared for physicians and jurists. Suggestive as these books are for disclosing different mental activities, they give no means of evaluating the disclosures. They show the puerile stage in diagnosis which had been reached before standardized tests were available.

Among those who are engaged in practical clinical work for determining mental development the Binet Scale has advocates who are quite as ardent as critics we have noted. Goddard, Kuhlmann (139), Wallin (213), and Towne (201), have all used it in the practical examination of hundreds of cases and heartily commend its use in connection with delinquents, as does Healy for the earlier ages (27, p. 80). On the other hand there is a growing sentiment that the examinations should only be entrusted to experts in mental development. It is felt that the physician who has not had enough training in a psychological laboratory to understand the snares of mental tests, and very few have had this opportunity, ought to refer this question to a clinical psychologist as the best physicians now do when such experts are available. Perhaps nobody is so well equipped to judge a child's mental development without diagnostic tests as his school teacher, although Terman has shown that the teacher's judgment may be seriously at fault when he has not learned to dissociate mental capacity from the age and size of the child (196). In an editorial in the Journal of Criminology, Dr. Gault (106, p. 322) expresses the opinion that “dissatisfaction with mental tests as a means of diagnosis” is traceable to the fact “that what the lay mind recognizes as palpable errors are often made by half-trained 'investigators,' 'research directors' and even by men and women whose only qualification is that they have been trained for six weeks in a psychological clinic.” Dr. Wallin demands that the tests should be used for diagnosis only by the psychologist with clinical experience.

The American Psychological Association has cautioned against diagnosis by those inadequately trained and adopted the following resolution at its 1915 meeting:

“Whereas, psychological diagnosis requires thorough technical training in all phases of mental testing, thorough acquaintance with the facts of mental development and with the various degrees of mental retardation.

“And whereas, there is evident tendency to appoint for this work persons whose training in clinical psychology and acquaintance with genetic and educational psychology are inadequate:

“Be it resolved, that this Association discourages the use of mental tests for practical psychological diagnosis by individuals psychologically unqualified for the work.”

Binet's suggestion as to the diagnosis of mental development seems to be best. He says that “the selection of defectives calls for three varieties of experience—that of teachers, of doctors, and of psychologists” (77, p. 38). These three points of view may be combined in a committee as in France, or the decision may rest with a specialist in mental development whose judgment should only be given after he has all the information which the medical, educational, and social diagnosis can provide to supplement his test records and his evaluation of the causes of the condition found.

Those who are considering the legal isolation of the feeble-minded, especially defective delinquents, and superintendents who wish a safe rule for transferring school children to special classes or schools for the mentally retarded should keep a committee plan in mind. A legal requirement embodying an examination by such a commission could easily be framed. In my opinion the expert in mental development should be required at least to have the equivalent of a year of graduate work with his major time in testing. On the other hand very desirable information as to children that require examination may be obtained by a teacher who uses a mental scale intelligently. In the hands of an amateur it may perform an analogous service to that of a vision chart in discovering children who require expert examination of their eyes. The danger lies in the novice not knowing his limitations. Few who have had experience with tests can doubt, however, the much greater danger of inadequate diagnosis of mental development on the part of physicians who give opinions about mental deficiency without having had experience with test scales.

A. In Minneapolis

Besides the estimates of deficiency based on tests, the school records may furnish valuable objective evidence about mental retardation among delinquents. The school environment is the first prominent social environment to which the child must adjust himself. If he fails in this while in regular attendance we have an important indication of mental deficiency. With laws which require attendance at school, we may even estimate the mental character of groups, on the basis of success in school, provided that we use proper caution as to the effects of late entrance and of absence from school. Moreover, whether retardation in school shows mental deficiency or not, it certainly sets forth a vital problem in connection with delinquency. We shall first consider the school retardation of delinquents and leave the problem of checking the tests by school records until later.

In order to study school retardation we tabulated the school position of 236 boys and 95 girls consecutively found delinquent in the Minneapolis juvenile court. To make the results more significant we did not include any cases dismissed at their hearing in court. Comparison with more serious delinquents is made by means of the group of 100 juvenile repeaters and 123 from the Glen Lake Farm School. The school position and actual age of each delinquent was compared with the age and grade distribution among Minneapolis elementary school children. The latter was determined by a census made the same year the returns for which included about 15,000 of each sex (see Table XII).^[31] The ages and grades were recorded for the beginning of September, when the school year opens, and the census was taken late in the year after all the children had been registered in school. That different groups can only be properly compared when the age-grade distributions are made for the same time in the year is clear when one remembers that the ages are changing throughout the school year while the grades remain the same for at least half the year. The census was taken for another purpose so that it unfortunately does not include the high school pupils. Since the frequency and amount of retardation increases for older ages which occur relatively more frequently in the groups of delinquents the comparison somewhat exaggerates the difference between the groups. This difference in the relative ages of the groups is allowed for, however, in a later table on which the discussion will be based. The school positions of the various groups of delinquents and of ordinary school children are given in Table XIII and graphically in Figure 2.

In the Minneapolis group of elementary school children it will be found that there is about as much chance of a child being in either of the two most common ages for a grade. Among the boys, for example, 36% were in the series represented by age 6 in the first grade, 7 in the second grade, 8 in the third grade, etc., while 30% were in the series represented by one year older for each grade. It is, therefore, reasonable to regard either 6 or 7 as a satisfactory age in the first grade, 7 or 8 in the second, when one estimates the amount of retardation in this group. The allowance of two ages as satisfactory for a grade is in conformity with the practise of Strayer (189). The necessity of taking these ages at either the beginning or the end of the school year, and not merely “in the grade,” is emphasized by the report of the New York City Committee on School Inquiry (72). Ayres (71) also considers only those pupils over-age who are over 7 in the first grade, 8 in the second, etc., so that this may be regarded as fairly well established as a standard for measuring the retardation in school position of groups of children.

The summary of results in Table XIII shows that 70% of the ordinary delinquent boys were retarded in school position as compared with 27% among the Minneapolis boys in the elementary schools, 91% of the ordinary delinquent girls as compared with 23% of the Minneapolis girls of these schools. When one compares the age distribution of the delinquent groups, given in Table XIII with that of the Minneapolis school children in Table XII, it is clear that an allowance should be made for the much larger proportion of older children in the delinquent groups. This may be done by determining the percentage retarded at each age and in each group and then calculating indices of retardation by weighting the percentage retarded at each age in the proportion to the number of delinquents at that age. Table XIV gives these results for the ages 8 to 15 inclusive.

For example, in calculating the indices 39 and 70 for the frequency of retardation among ordinary delinquent boys as compared with elementary school boys, the percentages retarded at each life-age for each of these groups was multiplied by the number of ordinary delinquent boys at this age, as shown lower in the table, and the totals divided by the number of ordinary delinquents, 213. The average frequency of the retardation of a school group which compares in ages with the delinquent group was thus determined. In calculating the indices of amount of retardation the same procedure is followed except that the average number of years retarded is found for each age and this is multiplied by the number of delinquents at that age. The 16-year-olds are omitted because of the inadequacy of the school census for this age. According to the standard which regards 7 years as satisfactory in the first grade there can be no retardation under eight years of age. Since some of the pupils 13 years of age and over have reached high school and so do not show in the Minneapolis table the percentage of retardation for children 13-15 years is based on the assumption that the number of children at these ages will be the same as the average number for 11 and 12 years. No credit could be allowed for those advanced in school positions on account of the incompleteness of the Minneapolis census for older ages. The comparison is, therefore, on the basis of retardation alone.