Many more estimates of the number of feeble-minded among school children might be cited, but they would add little to these authoritative samples. At the present time an estimate by health officers or teachers who are not familiar with the results of mental testing has little significance, as the whole complexion of the problem has been changed since the work of Binet and Simon.^[10] We may, however, cite three estimates based upon familiarity with test results, which fairly cover the range of estimates among school children. In connection with the Springfield, Illinois, survey conducted by the National Committee for Mental Hygiene under the direction of the Russel Sage Foundation, we find that three typical schools with a total of 924 pupils were studied. The report states that “the mentally defective children” constituted 3.8% of the number in attendance in March. The number of children in the schools examined, for whom instruction in special classes would be desirable, is about 7% of the entire enrollment of these schools (203, p. 10).

In connection with the Stanford Version of the Binet Scale, Dr. Lewis M. Terman says: “Whenever intelligence tests have been made in any considerable number in the schools, they have shown that not far from 2% of the children enrolled have a grade of intelligence which, however long they live, will never develop beyond the level which is normal to the average child of 11 or 12 years.... The more we learn about such children, the clearer it becomes that they must be looked upon as real defectives (57, p. 10). Again in placing the borderline for feeble-mindedness” with the Intelligence Quotient used, he suggests that “definite feeble-mindedness” lies below an I. Q. of 70 which with 1000 quotients was found to exclude about the lowest 1%. Above this is a group with I. Q.'s 70-80 which he describes as “borderline deficiency, sometimes classifiable as dullness, often as feeble-mindedness.” This group would include, as judged by the results of these tests, over 4% more.

Dr. Wallin, who has had wide experience in testing both school children and defectives, states: “I will venture the assertion, after years of teaching in the public schools and clinically examining public school cases, that the oft-repeated statement that 2% of the general school population is defective (if by this is meant feeble-minded), exaggerates the real situation. The actual number is probably about 1%” (211, p. 149).

After reading a paper on “A Percentage Definition of Intellectual Deficiency” before the American Psychological Association in 1915 (151), I was pleased to discover that Prof. Rudolf Pintner and Donald G. Paterson were also about to propose a percentage definition of feeble-mindedness for those who are dealing with mental tests (44). While their idea seems to be fundamentally similar, their paper shows that their conception is to be sharply distinguished in several particulars from that which I am advocating. They would limit the use of the term “feeble-mindedness” to individuals who test in a rather arbitrarily chosen lowest percentage of the population. As opposed to this I suggest continuing the present social definition of feeble-mindedness and supplementing it, for the purpose of aiding in the diagnosis, by indicating the social significance of those testing in certain lowest percentages. Such tested deficients I designate as “intellectually deficient.” It is important to consider their statement and to note what percentage they have chosen to regard as feeble-minded. They say:

“It is in order to avoid this vagueness and uncertainty attaching to the term that we suggest a definite psychological concept. The lowest three per cent. of the community at large, that is, the lowest as determined by definitely standardized mental tests, are to be called feeble-minded. Such a definition will be unambiguous and the dividing line between this and other groups will become clearer and clearer as we increase the accuracy of our measuring scales and the adequacy of our standardizations. Furthermore, if evolution is raising the degree of intelligence the three per cent. at the lower end will still remain, for, whatever the degree of their intelligence may be, they will still be feeble-minded as compared with the normal.

“Such a definition will in addition restrict the term to such as are lacking in intelligence and will differentiate them from the moral defectives and the psychopathic personalities, which are at present often confused with the group that we propose to call feeble-minded. An individual may be at the same time a moral defective and feeble-minded, but there is reason to believe that moral deficiency may exist without such intellectual defect as to warrant a diagnosis of feeble-mindedness. The same may be said of the psychopathic personality.

“The further question, whether all those coming within the proposed definition of feeble-mindedness are to be confined in institutions, is purely social and will be determined by the social needs of each community and does not concern us here. It is obvious that many more in addition to the feeble-minded as defined by us will require the restraint of an institution, even though no real mental defect exists.

“It is immaterial for the purposes of this hypothesis whether three or a smaller or larger percentage be designated as feeble-minded. The important point is the agreement upon some fixed percentage, and we have chosen three per cent. as covering presumably all the cases of marked mental deficiency. A brief glance at the chief estimates of the number of feeble-minded in civilized communities would indicate that our percentage is somewhat higher than the conservative writers give, but we shall show later on that it is much lower than the results obtained from groups of children tested by intelligence scales” (44, p. 36).

With those who understand that deficiency is mainly a question of degree, it would seem that there might be some agreement as to the plan for defining tested deficiency. In order to make this plan more useful to those dealing with the social care of the feeble-minded, it would be necessary to supplement the bare percentage definition by relating it to expectations of social failure somewhat after the manner I have attempted. In particular it will gain its main value for diagnostic purposes, it seems to me, if the percentage is so chosen that it may receive the support of conservative scientific opinion. To be most useful it seems evident, also, that the percentages must be chosen with regard to the sort of social care which it is anticipated would be justified for the particular degrees of deficiency.

Let us recall the percentages suggested to harmonize the estimates: the lowest 0.5% to be regarded as presumably deficient enough to justify isolation and the next 1% as doubtful, but low enough to warrant special training and probably requiring indefinite social assistance. If these percentages for tested intellectual deficiency have been shown to be fairly conservative estimates in the light of the authoritative judgments with which they have here been compared, the laboriousness of this comparison has been worth while. Further light upon the social assistance group may be thrown by the study of the success of those children who have already had the advantage of training in local classes for the deficient.

F. The Ability of the Mentally Retarded, Especially Those Receiving Special Training.

That we are not justified in isolating all whom we class as feeble-minded is best indicated by the evidence as to the number of these sent to special local classes for deficients who are able to float socially with the assistance of capable after-care committees. A fair picture of the present situation may be obtained by thinking of these pupils in the help-classes and schools as representing about the next 1% above those who have been isolated in institutions. With this picture in mind let us see what has been the outcome of their special instruction and social assistance thereafter.

In his book on Les Enfants Anormaux, Binet collected the evidence available at that time (77, p. 140). He says:

“Mme. Fuster, after a stay in Germany, where she visited some Hilfsschulen and Hilfsklassen (literally, 'help-schools' and 'help-classes') made a communication to the Société de l'Enfant, from which it appears that in the case of 90 classes for defectives in Berlin, 70% to 75% of the defective pupils who were there became able to carry on a trade; 20% to 30% died in the course of study, or returned to their homes, or were sent to medical institutions for idiots.

“According to a more recent inquiry, made under the auspices of M. de Gizycki at Berlin, and published in a book by Paul Dubois, 22% of the children were sent home or to asylums; 11% were apprenticed; 62% worked at occupations which required no knowledge and yielded little pay (laborers, crossing-sweepers, ragmen). If we add together these two last groups, we reach a proportion of 73% of defectives who have been made, or who have become more or less useful....

“Dr. Decroly has kindly arranged at our request a few figures relating to the occupational classification of the girls discharged from a special class in Brussels.... Finally, then, out of nineteen feeble-minded subjects, regarding whom particulars have been supplied, one-half, or 50%, have been apprenticed, or more than half, 75% if we count the defectives who 'work....'

“Through the intervention of an inspector, M. Belot, we have inquired of twenty heads of schools what has become of the defectives whom they notified to us two years ago. We have made these inquiries with regard to sixty-six children only.... If we subtract the two first groups, those about whom the particulars are wanting, and those who have not yet left school, there remain twenty-seven children, of whom seventeen have been apprenticed, or 76%.... Now this proportion is, by an unexpected agreement, identical with that obtained in the classes of Berlin and Brussels.”

A more recent report concerning the Hilfsschulen in Berlin by Rector Fuchs is in close agreement. It indicates that from 70% to 80% of the former pupils of these schools make a living after they leave school.

To compare with these reports indicating that about three-fourths of those leaving the special schools of Paris, Berlin and Brussels by social assistance attain occupational classifications, we have less favorable reports from Great Britain. Shuttleworth and Potts (181, p. 23) say:

“At the Conference of After-Care Committees held in Bristol on October 22, 1908, a paper read by Sir William Chance, Chairman of the National Association for the Feeble-Minded, dealing with the reports of the After-Care Committees of Birmingham, Bristol, Leicester, Liverpool, London, Northampton, Oldham and Plymouth. The combined statistics from the nine centers showed that 22% of those who had attended special schools for the mentally defective were in regular work, and 6.8% had irregular work.... To illustrate the necessity for continuous supervision and the futility of temporary care, we cannot do better than quote the records of the Birmingham After-Care Committee, as embodied in their report for 1908, after seven years work. It was found that, 'out of 308 feeble-minded persons who have left school and are still alive, only 19.8% are earning wages at all, and only 3.9% are earning as much as 10 s. per week'” (181).

Tredgold summarizes other data on this question of industrial success as follows:

“We may next turn to the reports of 'After-Care' Committees regarding feeble-minded (moron) pupils of the special schools. In London the proportion of pupils known to be in 'good or promising' employment was 37.5%. Two years previously it had been 45.7%, and Sir George Newman, the Chief Medical Officer to the Board of Education, attributes the falling off to two causes—firstly, insufficient after-care; and secondly, the two additional years. He remarks: 'The longer the test the more severe it is.' In Birmingham, the 'After-Care' Committee compiled information regarding 932 cases which had passed through the schools during the previous ten years. Of these, excluding the normal and dead, 272, or 34%, were engaged in remunerative work. At Liverpool, of 712 children passing through the hands of the 'After-Care' Committee during a period of six years, 85, or 11.9%, were doing remunerative work.

“Finally we may refer to some figures concerning 'After-Care' work compiled by Sir William Chance from the returns of the National Association for the Feeble-Minded. These were based upon an inquiry made of sixteen centers of the Association, and referred to a total of 3,283 persons. Of this number, 798 were doing remunerative work, 89 were 'doing work, but not reported;' 202 were useful at home; and 941 were returned as 'useless members of society.' If we exclude 340 who were transferred to normal schools (not being feeble-minded), we have 27% engaged in remunerative work.

“With regard to the term 'remunerative work,' however, it is to be remarked that the person employed is not being paid the standard wage. On the contrary, it is my experience that this is practically never the case, and this is corroborated by the observations of the secretary of the Birmingham center, who says: 'Although some of our cases have been at work for more than ten years, only 34 of the whole number (173) earn as much as 10 s., 2 d., per week. Of these only 6 earn as much as 15 s., and only 2 earn 20 s., which is the highest wages earned.... While it is not very difficult for some of our higher-grade cases to get work when they first leave school, it is almost impossible for them to retain their situations when they get older, and the difference between them and their fellows becomes accentuated. Uncontrolled and often quite improperly cared for, they rapidly deteriorate, the good results obtained by the training and discipline of the special school being under these circumstances distinctly evanescent.... There are few workers over twenty years of age'” (204, p. 425, 435).

The 1912 report of the London County Council (144) covers those who left its special schools for mentally defective children during the years 1908-1912 inclusive. These schools have accommodation for about 1% of the elementary school enrollment. Of 2010 children who left these schools during these five years, and who were still alive, 1357 were employed and 311 more employed when last heard from, a total of 79% employed at last accounts. Those out for five years show about the same proportion employed. This is a more favorable showing and fairly in line with the results of other European help-schools. The average weekly wages of those employed ranged from 4 s. 6 d. for those just out to 10 s. 10 d. for those leaving five years before. A considerable proportion who live at home thus have been meeting their necessary living expenses as the result of this special training and subsequent assistance.

Dr. Walter E. Fernald reported to the British Royal Commission on the Care and Control of the Feeble-Minded concerning the inmates of the institutions for feeble-minded in the United States. These institutions receive a much lower grade of cases on the whole than the local help-schools abroad: (83, Vol. VIII, p. 159)

“Some of the institutions where only the brightest class of imbeciles are received, and where the system of industrial training has been very carefully carried out, report that from 20% to 30% of the pupils are discharged as absolutely self-supporting. In other words at other institutions, where the lower grade cases are received, the percentage of cases so discharged is considerably less. It is safe to say that not over 10% to 15% of our inmates can be made self-supporting, in the sense of going out into the community and securing and retaining a situation, and prudently spending their earnings.... But it is safe to say that over 50% of the adults of the higher grade who have been under training from childhood are capable, under intelligent supervision, of doing a sufficient amount of work to pay for the actual cost of their support, whether in an institution or at home.”

The wages of the women at the Bedford Reformatory before entering prostitution as given by Davis (133, p. 210) have a direct bearing on the earning capacity of the higher grade feeble-minded. The Binet tests of Bedford women by Weidensall indicate that about 38% of the successive cases admitted to Bedford test in the lowest 0.5% intellectually, and 75% in the lowest 1.5% intellectually. Davis' table shows that for 110 whom she classes as mentally low grade cases at the reformatory, the median wage of those in domestic service, as claimed by the women, was nearly $4.50 before entering prostitution. These feeble-minded women, if their statements of earnings can be accepted, are therefore feeble-minded by reason of their low intelligence plus delinquency, and not by reason of inability to earn the necessities of life. The best of these mentally low grade cases earned as high as $5.00 in addition to board and lodging in domestic service and $25.00 outside of domestic service.

In this country we have fewer studies of the results of training the mentally retarded in special local classes and schools. Miss Farrell has made a preliminary report of 350 boys and girls out of the 600 children formerly in the ungraded classes in New York City during the preceding 8 years (102). Omitting seven whose status was unknown and 10 who had died, only 6% were known to have failed to survive socially with assistance. These were in penal or other institutions. On the other hand a strict analysis of her returns shows only 28% earning $5.00 a week or more and thus possibly surviving independently. Of the above group of 333, 86 were at home, 192 employed, 31 unemployed and 3 married.

In Detroit among 100 children over 16 years of age who had attended its special classes and been out of school not over 5 years, 27 had been arrested, but 39 of the boys had been at work and received an average wage of $7.00 per week, while 16 girls had averaged $3.75 in weekly wages, although few held their positions long (97).

Bronner (6) compared a random group of thirty delinquent women at the detention home maintained by the New York Probation Association with an intellectually similar group of 29 women all of whom had been earning their living in domestic service and none of whom had been “guilty of any known wrong doing.” The delinquents were 16 to 22 years of age while the servant group was somewhat older. Only two or three of the delinquent group were worse than the poorest of the servant group in any of the five intellectual tests, so that, if more than this number were intellectually deficient, they were no more deficient than those who had survived in society. No Binet scale records were published so that we have no means of determining how many of these delinquents might fall within either of our deficient groups.

The principal deduction from this evidence on the earning capacity of those of low intellectual grade is a caution against demanding the social isolation of all the intellectually weak until we have more definite information as to what portion of them are able to live moral lives, as well as earn their living with social assistance, without being cared for entirely in isolation colonies. That a significant number of the lowest 1.0% intellectually next above the lowest 0.5% have led moral lives and have shown considerable earning capacity after attending special schools, when they are given proper after-care, has probably been demonstrated. They should, therefore, be treated as an uncertain group whose feeble-mindedness would never be decided purely on the ground of the intellectual tests. Most of them will, however, probably be found mentally deficient enough to need at least social assistance and protection.

In concluding this summary on the estimates of the frequency of feeble-mindedness, it need only be added that so far as concerns the use of the percentage definition for fixing the borderline in any particular system of tests the percentages chosen are not essential to the plan. The principles of the method apply whatever percentages might be adopted. For such important purposes as the comparison of the relative frequency of deficiency in different social groups and harmonizing the investigations with different mental scales, agreement upon a particular percentage is not essential. In diagnosis, of course, it is a matter of fundamental importance in order that injustice may not be done individuals. For this reason the estimate should be conservative, possibly more conservative even than our tentative 0.5% at 15 years of age. Any investigator who disagrees with the above estimates of the degree of tested deficiency justifying isolation may substitute X per cent. with a doubtful region extending Y per cent. further. Provided such a census were legally authorized and funds available it would be not impossible to get a reliable determination by a house to house canvass showing the number of adult deficients, say 21 years of age, in typical communities, who were not able to survive socially without assistance. This number would then give the key for a conservative percentage and the movement for early care would be immensely advanced.

With the recent introduction of psychological tests into the cantonments of the national army, the goal of symptomatic borderlines as determined by objective tests seems to be almost at hand. Since the men are brought practically at random to the camps by the draft and are under military command, it may be possible to find out the social history of a large enough group at the lower limit of tested ability to establish the question of the necessary capacity for independent moral and social survival. These borderlines could then be transferred from the army tests to positions of equivalent difficulty in other test systems.

The remainder of this study will show some of the advantages of the percentage definition for fixing the borderlines with a system of tests and the result of applying such an interpretation to the particular problem of delinquency. The advantage in increased definiteness should already be evident. When a person is classed as presumably deficient it will mean that he is in the lowest 0.5% in intellectual development or within the lowest 1.5%, if he is a persistent delinquent.

Sufficiently large random groups have not been tested with any development scale to make the determination of the borderline on the scale more than tentative. Such borderlines must be looked upon as temporary descriptions to be used in aiding diagnosis until more data are available. Nevertheless, the percentage method of procedure seems to be an improvement over other plans of stating the borderline. So far as the Binet 1908 scale is concerned, when we supplement Goddard's results with 1500 school children by the data for the lower limits of a random group of 653 15-year-olds which we tested, the limits on the scale for passable intellects defined by the percentage method will be found, I believe, not only more conservative, but more reliable than those in current use. Moreover the intended meaning of such borders becomes clear.

A. The Border Region for the Mature.

(a) Indication from a random group.

The passing limit for adults is unquestionably much more important than that for children since any child who once passes this limit is assured, generally speaking, of social fitness so far as intellect is concerned. He has attained a position intellectually which is sufficiently good to enable him to get along without social assistance unless he is especially deficient in will. This borderline for the mature has been so thoroughly neglected that in none of the common published forms of the Binet scale, except the new Stanford Scale, is there an attempt to define it. This seems almost incredible in view of the general use of the Binet method in diagnosing feeble-mindedness. To be sure, there are discussions of this upper limit, as we shall see, but they have usually not been embodied in the actual directions accompanying the scales which get into the hands of amateurs. Most of these directions content themselves with describing borderlines for children with no caution about the final lower limit for social survival.

The borderline for the mature is the first difficulty which a court examiner will encounter when he attempts to obtain assistance from an objective system of measurement. Very little experience will convince one that it is not enough to describe the deficient ability of an adult in terms of years of retardation. It is widely agreed that at some age during adolescence practically all the mental processes are available that will be found in the mature. From that time the advance in ability is made by attaining greater skill in specific activities through training and by increasing knowledge, rather than through a native change in the form of thinking. If mental tests mainly reach capacity for thinking, as they aim to do, rather than amount of knowledge or skill in specific work, then we are conservative in using a randomly selected group at 15 years of age for approximating the borderline on the scale for the mature.

In connection with the new Stanford Scale, Terman says: “Native intelligence, in so far as it can be measured by tests now available, appears to improve but little after the age of 15 or 16 years. It follows that in calculating the I Q (intelligence quotient) of an adult subject, it will be necessary to disregard the years he has lived beyond the point where intelligence attains its final development. Although the location of this point is not exactly known, it will be sufficiently accurate for our purpose to assume its location at 16 years” (57, p. 140).

Yerkes and Bridges in connection with their Point Scale say, “it seems highly probable that the adult level is attained as early as the sixteenth year” (225, p. 64). Kuhlmann (138) used 15 years as the divisor in calculating the intelligence quotient of adults and Spearman thinks that the limit of native development is reached about 15 years (184). He says, “That mental ability reaches its full development about the period of puberty is still further evidenced by physiology. For the human brain has been shown to attain its maximum weight between the ages of 10 and 15 years” (184). For the last statement he quotes Vierordt. On the contrary Wallin thinks that we need more evidence for the correctness of these hypotheses before choosing a fixed age as a divisor for adults (215, p. 67).

We are not interested in determining a divisor for an adult intelligence quotient but in fixing a conservative borderline for the mature. Admitting that the mental capacity of those 15-year-olds at the lower limit may not be like adults, nevertheless adults would be more likely to be better than worse. Borderlines for the 15-year-olds, should, therefore, be safe for adults. Moreover, the lower limits with a truly random group of 15-year-olds would probably be more reliable than an assorted group of adults subjectively chosen from different walks in life and combined in an effort to represent a random mature group. The Stanford Scale utilizes such combination of selected adults. It seems, therefore, that we are justified in utilizing the lowest percentages of randomly selected 15-year-olds as a reasonable criterion for describing the limits for adult deficiency. Surely adults below this lower limit for 15-year-olds would have questionable intellectual capacity.

The borderline for the mature being the crucial feature of a developmental scale when used for detecting feeble-mindedness, it seemed imperative to us that some effort should be made to obtain records with a random group of older-age children or adults. Goddard's results with school children were not significant above eleven years of age since the personal examinations were confined to children in the sixth grade or below. The twelve year old group in the sixth grade clearly omits the best 12-year-olds, so that the percentage method would have no significance applied to his figures for children above 11 years of age. Moreover it was obvious that the group of public school children 15 years of age or older would not give a picture of the lower end of a random group since many children drop out of school at 14. On the average those that leave are undoubtedly of lower ability than those who remain.

The most valuable data on the borderline for the mature would come from mental examinations of large random groups of adults. The impossibility of gaining the consent of adults for such examinations puts this plan out of consideration. Perhaps the next best method would be to examine all the children of 15 and 16 years of age in typical communities. It happened that we could approach this result in Minneapolis since we there had an excellent school census made from house to house covering all children under 16 years of age. The Minnesota law requires school attendance until 16 years of age unless the child has graduated from the eighth grade. Under the able direction of Mr. D. H. Holbrook of the attendance department the census of children of school age had been made with unusual care. All the children living in each elementary school district in the city were listed in a card index regardless of whether they were attending public, parochial or private schools, or had been excused from attendance for disability or for any other reason. Since we only needed to be sure to examine the lowest few per cent. of the children in ability this group of 15-year-olds could be tested by examining all those children in typical school districts in the city who had not graduated from the eighth grade. A third of the 15-year-olds were still in the eighth grade or below. Neither the compulsory attendance law nor the census would have reached the 16-year-old adequately. In most states even the 15-year-olds would have been above the compulsory school age.

There were 653 children, (322 boys,) 15 years of age living in the seven typical districts which were selected objectively for study. Among these there were 196 who had not graduated from the eighth grade. All of these latter children were examined, except one who could not be tested as she was in a hospital on account of illness. Quite a number of the children were in parochial or private schools, two were followed to the state industrial school and a number were examined at home. In order to be sure that we had not missed any institutional cases in these districts the complete list of Minneapolis children at the State School for Feeble-Minded was gone through to get any of low ability who might have been missed.

The seven districts in which the children were to be studied were chosen, with the idea of avoiding any personal bias in their selection, by taking them alphabetically by the name of the schools, except that no district was taken where the normal school attendance of the district was affected by inadequate school facilities so that children had to be transferred either to or from that district to other schools in order to meet crowded conditions. It happened fortunately that none of these schools represented extreme conditions in the city. The average percentage of children in the 69 elementary schools of the city retarded in school position below a standard of 7 years in the first grade, 8 in the second, etc., was 24.1% with a mean variation of 6.5%. The percentages retarded in the schools studied were as follows: Adams, 22.7; Bryant, 21.1; Calhoun, 21.7; Corcoran, 29.4; Douglas, 20.4; Garfield, 18.6; Greeley, 26.4.

Kuhlmann's adaptation of the 1911 scale (135) was used as a basis for the examinations, supplemented by the 1908 scale wherever tests had been changed so that other forms of the tests were found in either Kuhlmann's (136) or Goddard's (110) adaptations of the 1908 scale. Since test results with the 1908 scale provide the most data for describing the borderline for the immature, our plan was to use the 1908 form of a test first when the procedure had changed. The supplementary directions were arranged for each age so that the testing could proceed methodically and the results be scored under either the 1908 or 1911 scale with the least possible disturbance of each test. Over a third of the children were tested by myself. The rest were tested by three advanced students in psychology. It is a pleasure to express my thanks to these assistants, Miss Rita McMullan, Miss Lucile Newcomb and Miss Florence Wells. Besides having had brief experience in dealing with exceptional children, they practised testing under my observation until the tests could be given smoothly and I was convinced of their ability to follow directions intelligently and make full records with reasonable accuracy. The results of the tests were all carefully gone over and scored by me. So far as I can judge, the results are quite as accurate as any other published tables, although one must always consider the possible effect of errors of testing. Separate rooms were provided at the schools or homes so that the child could be alone with the examiner during the testing.

In attempting to define the borderlines on these scales we might either state the exact scale position in tenths of a year below which 0.5 and 1.5% of the cases fall, or we might merely attempt at present to state the borderlines in rounded terms of years on the scale. The latter plan is the one I have adopted for several reasons. The main reason is that I wish to emphasize that these are still rough boundaries. Besides that, however, a study of the results shows that the cases do not distribute by separate tenths of a year so that exactly these percentages could be picked off, without a questionable smoothing of the curves while the rounded years approach these limits fairly well.

It seems to me that it is best at present to be carefully conservative in describing these borderlines, so that I have chosen them from the available data at the nearest rounded age position which is reasonably sure not to catch more than these limiting percentages. Throughout the tables I have also followed the published directions for the 1908 scale in classing the person in the intellectual age group in which he finally scores all or all but one of the tests. I recognize, of course, that this is an arbitrary limit; but it is the limit fixed by the usual printed directions going with the 1908 scale, which is the only one thus far standardized for the immature on the percentage basis. For those who wish to calculate other borderlines or reconstruct the individual tests of the scale I have provided the complete data for each individual both for the 1908 and 1911 scales in Table XXI, Appendix I. The table also gives the exact ages and school grades of each child.

The summary of the results with the tests for those testing under XII is given in Table III. Life-age^[11] at the last birthday and not the nearest life-age is used in the table. The children were all between their 15th and 16th birthdays. Following the directions published with the scales, the basal age for calculating the results in the table is taken as the highest at which all or all but one test are passed for the 1908 scale, and the highest at which all were passed for the 1911 scale. Two-tenths is allowed in the table for each test passed above the basal age and 0.1 for an uncertain answer. The children were tested by the long method, beginning with the mental-age group at which the child could pass all the tests and continuing to that age group in which he failed in all.

Thrown into percentages of the group of 653 children living in these districts, it is evident that a test score of XI raises any person above the group of intellectual deficients. The percentage that tested this low, i. e., under XI.8, with the 1908 scale, was 10.0 (65 cases) and this would probably be increased if those who had graduated from the eighth grade had also been tested. The percentage testing under the same position in the 1911 scale is 8.1 (53 cases). With the 1911 scale there were 32 additional cases testing XI.8 or XI.9. The table indicates that 0.2% of the 15-year-olds tested below IX.8 with the 1908 scale, and 0.5% with the 1911 scale. This defines our scale borderline for the mature who are presumably deficient as below test-age X. These positions are near enough to the lowest 0.5%. The group testing of uncertain ability, age X, (strictly speaking between IX.8 and X.7 inclusive,) includes 0.7 to 0.9%. We thus approach fairly well the rounded age positions which exclude 1.0% above the lowest 0.5%. The total number testing in presumably and uncertain groups is thus 1.1%, 7 cases out of 653, for the 1908 scale and 1.2%, 8 cases, for the 1911 scale. This is to be compared with the percentage definition that the lowest 1.5% are either presumably deficient or uncertain.

At present we are entitled to assume that adults testing below XI, i. e., below X.8, are so low in intellectual development that it is a question whether they have sufficient equipment to survive socially. Fine discriminations with the Binet scale are not possible with our present knowledge. So far as our information goes, if we use the percentage method of defining intellectual deficiency, we may say that adults who test X are in an uncertain group in intellectual ability, with the probability that they will require more or less social care, while those who test IX are deficient enough to need continuous care unless the evidence of the test is contradicted by other facts or is accounted for by the existence of removable handicaps.

It is perhaps not necessary to call attention to the fact that X and XI are used here merely to refer to positions on the Binet scale without regard to what per cent. of ordinary 10-and-11-year-old children attain these positions. For example, XI does not imply that most of the children of eleven years of age are above this borderline. Table IV, to be given later, suggests that hardly two-thirds of random 12-year-old children pass this position on the 1908 scale and not half of the 11-year-olds. Thorndike regarded X.8 as normal for a child of 11.6 years of age. (200)

So far as the determination of intellectual deficiency is concerned we should note with emphasis that placing the limit of passable intellects at XI for adults almost entirely removes the common objection to the Binet scale on account of the difficulty of the older age tests. The older age tests become of little consequence because the best of the deficient group have a chance at tests in at least two groups above those of mental age X, so that they can increase their score by passing advanced tests as they could not if they had to test XII.

As a check upon the borderline for those presumably deficient, it is important to note that the only case which tested below this borderline with the 1908 scale was a girl in the 4B grade. She tested exactly IX with each scale and was the only child in the group who was below the fifth grade in school. There can be no question that she was mentally deficient. On the other hand in the group which tested X or above there are several cases which it would be unjust in my opinion to send to an institution for the feeble-minded without some other evidence of mental weakness. Half of them, for example, are in the seventh grade. In Minneapolis this is not as significant as it might be in other cities, since pupils are rarely allowed to remain more than two years in the same grade whether they are able to carry the work of the next higher grade or not. Pupils in higher grades may not always be able to do even fifth grade work.

The evidence from the institutions for the feeble-minded indicates that less than 5% of their inmates test XI or over. Of 1266 examinations at the Minnesota School for Feeble-Minded, 3.8% (154); of 378 examined at Vineland, 3.2% (113); of 140 consecutive admissions examined by Huey at Illinois, 5.7% (129). To be sure, a goodly number of these inmates are not eleven years of age, but a majority of them are at least that old and many are older. Of 280 children in the Breslau Hilfsschulen, Chotzen (89) found none reaching XI, and only six who tested X. These few cases in institutions reaching XI or over may well come within our class of those feeble-minded through volitional deficiency.

Goddard's description of the children at the Vineland school for feeble-minded who tested XI with the 1908 scale hardly sounds like an account of social deficiency. He says:

“In the eleven year old group we find only five individuals, but they are children who, for example, can care for the supervisor's room entirely, can take care of animals entirely satisfactorily, and who require little or no supervision. They are, it is true, not quite as expert or trustworthy as those a year older, and yet the difference is very little and the two ages can probably be very well classed together” (113).

The studies of groups are more important for fixing our general rules than individual examples. We must always expect to find exceptional cases where the brief intellectual tests given in an hour or less are not adequate, especially if the testing has been interfered with by the person's emotional condition at the time or by deliberate deception. A number of illustrations have been reported of successful adults who have tested X under careful examinations. Such, for example, are three cases of successful farmers tested by Wallin (215) and a normal school student tested by Weidensall (59). There are two examples of persons testing IX with the Binet scale and yet earning a living. Such is the case related by Dr. Glueck of the Italian immigrant making two trips to this country to accumulate wealth for his family by his labor (109), and the case of the boy reported by Miss Schmidt (179). These cases should make us cautious, but they are so rare that it seems best to treat those testing IX at least as exceptions.

The group studies confirm our suggestion that a borderline of X or below will bring in for expert consideration nearly all adults who are feeble-minded from a lack of intellectual ability, while testing IX is a fairly clear indication of such serious deficiency as to justify isolation. That testing X, in the absence of other evidence of conative disturbance, places the case only in an uncertain region so far as isolation is concerned is best indicated by the fact that 1.1% to 1.4% of these 15-year-olds tested this low. We have good evidence that many in special classes, which contain only about the lowest one per cent., afterwards do float in society with or without social assistance. They cannot be presumed to require isolation, as I showed in the previous chapter. It is better to say at present that those testing X require evidence of their deficiency before isolation, except in special classes, is justified. The test diagnosis alone is too uncertain, even when there are no removable handicaps.

As to the reliability of these borderlines, too much emphasis can hardly be put upon the fact that they have been determined for only a single group of 653 in a single community. They are undoubtedly not the exact borderlines, although they are the most probable percentage estimates we have at present and were obtained in a group that was as nearly unselected as it is possible to obtain. The method of selection was perfectly objective and excluded no feeble-minded children of this age living in these school districts.

The theory of sampling applied to percentages (228) enables us to say that the standard deviation of the true lowest 0.5% in samples of this size made under the same conditions would not be more than 0.28%.^[12] That is to say, if our result were only affected by the size of our sample the chances are about two out of three that the border of the true lowest 0.5 per cent. would lie between the border of the lowest 0.22% and the lowest 0.78% of a very large sample. Assuming that the distribution in this sample represented that of communities generally, the chances would be two out of three that the true border of the lowest 0.5% for like groups in like communities examined under the same conditions would lie between IX.0 and X.6 or X.4 on the 1908 and 1911 scales respectively. Moreover, the chances that a case in the lowest 0.5% in this sample would be above the doubtful group in a larger sample, i. e., get above the lowest 1.5%, would be about 1 in 10,000. On the other hand, the chances that a case above the true lowest 1.5%, i. e., above the uncertain group, would get into the lowest 0.5% in a larger sample, i. e., be classed as clearly deficient intellectually, would be about 18 in 1,000.

So far as the theory of sampling goes it would seem that these borderlines for the mature are sufficiently accurate for correcting present practise. On the other hand, the conditions in Minneapolis so far as deficiency is concerned are probably better than in the country as a whole, so that the borderlines here described might very well exclude more than the lowest 0.5% and 1.5% in the country at large. But if we shifted the definition so as to exclude the lowest 0.2% and 1.1% (the percentages empirically found below the limits described), the borders on the Binet 1908 scale would not be changed from the rough measures IX and X which are as accurate as we should expect to define our limits with the present data.

(b) The Present Tendency Among Examiners.

Comparing the suggestions as to the borderline for the mature which have heretofore been made, we find that they have gradually approached the boundary now suggested by the percentage method. In 1910 the American Association for the Study of the Feeble-Minded adopted a tentative classification in which the upper limit of the feeble-minded included those “whose mental development does not exceed that of a child of about twelve years” (64). This was based mainly on the fact that Goddard had found no case at the Vineland school for feeble-minded which tested higher than XII. Huey later than this found only two such cases at the institution at Lincoln, Ill., and Kuhlmann only ten cases at the Minnesota State School for the Feeble-Minded.

There was an early statement by Binet which referred to the practise in Belgium of regarding older school children as deficient when they were three years retarded in their school work (77, p. 41). This practise may have also contributed to this formulation by the American Association. Binet, however, regarded a child of the mentality of twelve as normal. In 1905, before his tests were arranged in age groups, he said:

“Lastly we have noticed that children of twelve years can mostly reply to abstract questions. Provisionally we limit mental development at this point. A moron shows himself by his inability to handle verbal abstractions; he does not understand them sufficiently to reply satisfactorily” (76, p. 146).

It is important to consider how the suggestion of XII as the upper limit of feeble-mindedness for adults got into the early practise in this country as the lower borderline for the mature. It is the most serious error which has marred investigations in this field. It seems to have been a case of repeated misunderstanding on the part of examiners for which nobody in particular was to blame. So far as I can determine nobody stated directly in connection with any scale what should be regarded as the lower borderline for the mature. Numerous examiners, however, in reporting their results, concluded that if the feeble-minded tested as high as XII then adults who tested XII were feeble-minded. They were somewhat encouraged in this fallacy by the fact that the 1908 scales suggested three years of retardation as an indication of feeble-mindedness, and the highest age-group of tests was soon shifted to fifteen years.

The trouble seems to have been that early workers failed to recognize that some of the feeble-minded in institutions, the purely conative cases, have passable capacity so far as the brief intellectual tests are concerned. To determine scientifically what is the borderline, we should study randomly selected groups from the general population and determine the positions on the scale below which practically all are socially unfit. Or, as Wallin has suggested, we should find out the degree of tested ability necessary for survival in simple occupations that are afforded by society (216, p. 224). These positions can only be checked by finding the conditions in institutions or special classes. They cannot be determined by tests of these abnormal groups alone. Besides the confusion arising from these feeble-minded who are primarily unstable or inert, but with passable intellects, reasoning from the statistics on abnormal groups merely repeats a common fallacy. The fact that some inmates of institutions test XII does not let us know how many outside the institutions who test XII actually survive in society.

The randomly selected groups of children on which Binet tried out his tests were so ridiculously small that he continually cautioned against adopting his suggestions as to borderlines as anything but tentative. For judging the borderline for the mature there were no test results which had not been seriously affected by the methods of selecting the groups, so we collected the data on this random group of Minneapolis 15-year-olds. I trust that this will make any examiner more careful about assuming that adults testing XI are clearly unable to survive socially, unless he is ready to claim that 10% of the general population are unfit socially.

It is to be noted that, taken literally, the description of the American Association is not in terms of the Binet scale, but of the mental development of a normal child of twelve years, although the framers of the resolution undoubtedly had the Binet scale of mental ages in mind. It was soon found that the tests for the older ages in the Binet 1908 scale were too difficult for the places assigned them. This is certainly true with the tests for twelve years and probably with those for eleven. This evidence is assembled in Table IV. The combined results should be used only with great caution since the methods of the investigators differed in detail and the groups were differently chosen. In the groups of children which Bobertag and Bloch and Preiss tested, there had been eliminated some of those who were backward in school, while Goddard's group did not include the best 12-year-olds.