been made of group intelligence tests in the St. Louis school system, but the school work done in this system admittedly ranks with the best in the country. It would be truer to say that children can not be classified or promoted in their school work solely on the basis of their "general intelligence"

for the right place," that the scale assures parents that their children are receiving "just the sort of instruction which the schools should give them," and that the use of the scale will reduce retardation to "a negligible quantity" as if the giving of the tests would transform an inherent dullard into a capable pupil and dispense with the necessity for repetition! An enormous amount of repetition is needed even when subnormals are placed in special classes. The following claim is made for another intelligence scale. "Will find your bright children during the first twenty minutes of the first day at school and will show in the same time the relative intelligence rating of all your children of any age and grade." Certainly a group intelligence scale would aid the educator in the accomplishment of this desirable goal. But it is misleading to assume that all the pupils would be correctly placed by the scale.

It has been frequently stated that any one who can teach school can give group intelligence tests. This is undoubtedly true so far as concerns the automatic phases of testing. But there is more to group testing than the automatic administration and scoring of the tests. Even this can not be well done without automatizing the formal procedure. But even the group tester must be skilled in the non-automatic phases of group testing: she must be able to inspire, secure willing cooperation and sustained application, she must make relevant observations of the peculiar reaction types of various pupils to secure a basis for interpretation of the scores, etc. I recently observed a child perform in a group test given in another. school system whose reaction type may not be so unusual. He did everything accurately which he was told to do by the examiner in all of the tests within his range of ability, but he failed every time to make the lead pencil check when the examiner called "stop." He scored zero in the examination, but the examiner paid no attention to the pupil's mode of reaction, and obviously could not explain the result when the individual examination showed that the child was merely backward.

as determined by a group intelligence examination, or even an individual intelligence examination,22 but that the results of such examinations should always be utilized whenever it is possible to do so. A satisfactory pedagogical classification must take into account not only the child's "general intelligence," but also his specific pedagogical, intellectual, volitional, emotional and physical idiosyncrasies, special abilities and disabilities, together with the significant genetic or longitudinal factors in his development (and to do this requires, of course, a psychoclinical approach). It is necessary to emphasize this point because there is apparent to-day the same tendency to exaggerate the uses and possibilities of group intelligence tests which was apparent five or ten years ago in the case of the Binet tests.23 Part of this exaggeration is, no doubt, to be ascribed to clever advertisers who want to dispose of their wares. There is now no necessity for attempting to popularize group intelligence testing by propaganda of exaggera

22 This conclusion has been stressed before: "The Achievement of Mental Defectives in Standardized Educational Tests," SCHOOL AND SOCIETY, 1919, 250 f. "The Achievement of Subnormal Children in Standardized Educational Tests" (a monograph delayed in publication).

23 As an illustration of this type of exaggeration even when applied to individual examinations let us cite the following statement, published in a northwest daily on June 13, 1920, under the signature of a professor of psychology in a state university: "The problem of how by an hour's examination we can put the man on the job for which he is fitted, measure his skill for that job as accurately as we can measure the depth of a river, and tell whether the applicant must work as an apprentice under constant supervision, or as a journeyman with occasional supervision, or as an expert capable of being a boss, has been solved by the personnel staff of the United States army." Exaggerated claims are eventually exposed, and in the final analysis only retard the cause which they aim to promote.

tion. Rather the need now is for critical evaluation and caution-caution against the making of unwarranted claims and against arousing expectations that can not be fulfilled.

Let us not only admit but emphasize that group intelligence tests are of value if used as one among many aids in educational, intelligence and social classifications. But they are only one aid; they do not enable us to dispense with other relevant facts; they supplement but certainly do not serve as a satisfactory substitute for the pedagogical data (obtained from the school record); and the responses and scores obtained in group intelligence tests always call for analytic interpretation. Where no pedagogical record is available, or where a satisfactory estimate of the child can not be made because of a defective pedagogical record, the group intelligence tests naturally will be of considerable value. Beyond this two specific values of the group tests appeared clearly in this experiment in the St. Louis schools. First, in one school we culled out the mental defectives earlier than we should have done by our regular method. The principal explained that he recognized that the pupils were abnormal, but he wanted to allow them more time before reporting them for examination. There are, of course, obvious advantages in transferring very young mental defectives to special schools, but in practise we find that many parents will withhold consent until the child has had a thorough trial in the regular schools, while it takes years to reach a positive diagnosis in many cases. Second, the testing proved to be very popular in the schools, due in part to the widespread popularity of group testing since the war. The group testing afforded the schools an opportunity to see the clinic at work. It afforded a more direct con

tact with the schools. Some school people do not realize the amount of work that goes on "behind the scenes" in a psychoeducational clinic.

If group intelligence tests only enable us to diagnose, classify and place individuals somewhat better than we are able to do without them they will serve a useful purpose, and will be worth the time and trouble involved in using them. But they must always be used with judgment and an appreciation of their inherent limitation.

3. Group intelligence tests will not take the place of psycho-clinical examinations, for various reasons, among others the following:

First, because an adequate picture even of the individual's intellectual peculiarities can not be obtained from a group intelligence test. To get such a picture from a group intelligence scale would require the use of so many tests as to make the scale too cumbersome. In the individual examination it is possible to give scores of formal intelligence tests designed to reveal specific defects and talents, while it is also possible to make many observations regarding the individual's intellectual, volitional and emotional peculiarities. A fairly detailed psychogram can be obtained from an analysis of the responses in the Binet scale alone, owing to the considerable number and variety of the tests which it contains. The clinical picture can, of course, be made more complete by giving other intelligence tests and various psycho-motor, anthropometric, physical and pedagogical tests, according to the time available and the exigencies of the case, supplemented by relevant facts from the personal and family history.

Second, the results from the group tests are occasionally quite unreliable and misleading. We have already given instances

of marked discrepancy between the Pressey score and the diagnosis made as a result of the clinical examination. We have occasionally found equally notorious discrepancies when other group tests have been employed. To cite merely one instance:

J. R. was reported on Form 13-H from an elementary school at the age of 13-4. The examination was not requested because it seemed evident from the record (he did successful work in the fourth grade) that he belonged in an ungraded class and not in a special school for the feeble-minded. Seven months later he entered a junior high school, but was not given a group intelligence test until he had been in the school 8 months. In this test he graded 9 years 9 months, and made an I.Q. of .68 and a coefficient of brightness of .70 and, because of the low score, was immediately reported to the clinic as a "low moron" (according to the automatic standards of diagnosis promulgated to enable any teacher to diagnose) who could not be retained in the school because he was feeble-minded and therefore not responsible for his deportment, which was not the best, and who should be in a school for the feeble-minded. The clinical examination showed that he was not feebleminded at all. At the age of 14-7 he graded 13-2 by the Stanford-Binet, thus being retarded only 1-2, and having an I.Q. of .90 (or an I.Q. of 100 according to the army adult norm). His basal age was X. He passed tests as high as sixteen.

Although subnormal in physical development, he appeared to be considerably older than he was, his state of nutrition was poor, he had deepset myopic eyes, a hoarse voice, was diagnosed as a case of neurotic inadequacy of slightly subnormal intelligence, and was recommended to an ungraded class in place of an elementary

industrial school, which was not available. In this case the low score in the group intelligence test engendered an attitude toward the boy on the part of the junior high school staff which was distinctly pejudicial to his best interests. It is unfortunate that in most cases where discrepancies occur (at least in our experience) between the group and individual tests, the group rating is too low. Moreover, however accurate and valuable group tests may be with the average run of cases, it would be hazardous to make clinical diagnoses of major consequence on the basis of group scores unless the tests indicated the presence of capacities, or unless the results were in complete harmony with, and could be consistently and satisfactorily interpreted by other relevant facts.

What, then, may we hope to get, at best, from standardized group tests or scales of intelligence? We can get (1) fairly accurate measures of central tendencies, (2) measures of variation from such tendencies, (3) measures of correlation between various tests and various determinate factors, and (4) the location or classification of an individual with reference to the measure of central tendency, and of the measure of variability when the tests furnish a basis for such computation. But the measure of central tendency is largely a mathematical fiction representing no particular individual and no particular type, but merely the mean of various types some of which vary in opposite direction from the postulated average or normal person, while measures of correlation between traits also merely represent the mean of contrasting types. The study of individual differences in psychology has shown that individuals range themselves into various types, that there are variations even within the recognizable

types, and that the principles of mental behavior which apply to one type may not apply to another. Obviously, the principles which are applicable to the "average" may not exactly fit the needs of any particular individual. In the last analysis the rightness or wrongness of any particular method of instruction depends on its power to satisfy the psychological needs of the individual concerned, rather than the hypothetical average, and, ultimately, it is with particular cases that the educator, the clinical psychologist, and the physician has to deal. Group intelligence tests and results can give us valuable points of reference-average or standard scores, or distribution curves but after the individual has been properly placed in the scale we must still try to discover the special variations found within his type-i.e., in what respects he is exceptional for his own type-the causes of such variations, and the differential treatment which will meet his particular needs. To ascertain, even imperfectly, such facts as these requires an analytical or clinical study of each case.

Naturally the clinical psychologist welcomes the results of group intelligence tests, as he welcomes all items of information from whatever source which may have a bearing upon his case. His chief task is to critically interpret all his findings synoptically, to sift the relevant from the irrelevant factors, to correctly evaluate the rôle of the different factors from the standpoint of etiology, treatment and prognosis. Nothing less will afford a rounded or adequate clinical picture. The ideal at the best can, of course, only be imperfectly realized in our present state of knowledge. J. E. WALLACE WALLIN ST. LOUIS PSYCHO EDUCATIONAL CLINIC AND SPECIAL SCHOOLS

THE SCHOOL SURVEY FROM THE SUPERINTENDENT'S POINT OF VIEW1

I SHALL base my remarks upon a Biblical text, which is found in Exodus, 3: 10-11; 4:1-3:

Come now therefore, and I will send thee unto Pharaoh, that thou mayest bring forth my people the children of Israel out of Egypt.

And Moses said unto God, Who am I, that I should go unto Pharaoh, and that I should bring forth the children of Israel out of Egypt? ... behold, they will not believe me, nor hearken unto my voice: for they will say, The Lord hath not appeared unto thee.

And the Lord said unto him, What is that in thine hand? And he said, A rod.

And he said, Cast it on the ground. And he cast it on the ground, and it became a serpent.

The point of this narrative is that here was a man who was asked to do a great work, and he made the excuse that he could not do what was required of him because he had no equipment. He was not eloquent; he had no money, and no armies, and no influence. How could a plain shepherd do this great thing? He was told to begin with what he had. It was nothing but a shepherd's staff. But by means of this he performed his miracles and achieved God's purposes.

1 On December 10, 1920, at Washington Irving High School, the New York Society for the Experimental Study of Education discussed "Mental Surveys in Public Schools." The following was the program:

1. A Mental Survey of Public School 11, Manhattan, by Professor Leta S. Hollingworth, Teachers' College, Columbia University.

2. Individual Case Work in Public School 11, by Miss Eleanor Hope Johnson.

3. Reorganization on the Basis of the Mental Survey, by Principal Benjamin B. Greenberg, Public School 11.

4. The Social Importance of Preventing Childhood Maladjustments, by Dr. Sanger Brown, 2nd, Neurological Institute.

5. The Mental Survey from the Superintendent's Point of View, by District Superintendent Joseph S. Taylor, in charge of Public School 11.

I was interested in surveys twenty-five years ago, when they were known as child study. In the fifth volume of Pedagogical Seminary, published in 1898, you will find a study of mine on Interest. Two thousand children of this city and of Winona, Minnesota, were asked what subject in the curriculum they particularly disliked, and why, and what subject they liked better than anything else, and why. I discovered some things worth while. I found out that arithmetic at that time was the subject of overshadowing influence. It was preferred by more children than any other subject. It was also hated more than any other subject. Forty-two per cent. liked it, and fourteen per cent. despised it. History came next in popularity. Nature study, music, writing, and drawing were at the foot of the list of likes. My interpretation was then and is now that if you wish to locate poor teaching find the subjects that children dislike. Our teaching of nature study, music, and drawing in those days was very crude, and the children so characterized it by their votes. Our teaching of arithmetic was better then than it is now, and the children so testified.

I, also, a quarter of a century ago, made a health survey of every class I taught. This was long before there were school doctors and school nurses in New York, and long before my official superiors paid any attention to such matters. I tested the sight and hearing of all my boys and systematically recorded their heights and weights. The figures of Porter and Bowditch were available in those days, and I had taken a course in anthropometry from Professor Burnham at Clark University. In other words, at the beginning of every term I answered the question, to the best of my ability: "What is that in thine hand?"

It was by such means that I discovered that the delinquent child in school is nearly always the child with some physical or mental defect or with an unfavorable social environment. By removing the cause you remove the trouble. That is why I am asking, but asking in vain, in my present position for visiting teachers in exchange for an equal number of attendance officers. Judge Lindsay gives expression to the same idea in his clinical laboratory, which is attached to his court. He never makes a final decision on a juvenile delinquent until he has ascertained the three ages of the child-the chronological age, the mental age, and the physical age. A girl with a chronological age of 14, a physical age of 18, and a mental age of 10 is almost certain to get into trouble. Hence the court is not meting out punishment, but is trying to find the best means of adjusting the delinquent to his environment.

One of the problems we have set for ourselves in my district this year is the reduction of retardation. The first question we asked was, "What is that in thine hand?" In other words, what are the present facts of retardation in the district? I called upon Mr. Nifenecker, the Director of our Bureau of Reference and Research, for help; and he very kindly took our last age-grade tables and worked them up into percentages for each school and grade. We are now studying causes and remedies of retardation and hope to make considerable progress in this vital matter before the year is out.

When a superintendent finds a subject below grade in his district or in one of his schools, he is frequently met with an excuse like this: "You must remember that these children are Italians. You can get results in the Bronx, among the Hebrews, but these children are different." In some cases my