JV3\ , 5, Ml Whole No. 1 Series 1, No. 1 HARVARD MONOGRAPHS IN EDUCATION A COMPARISON OF THE INTELLIGENCE AND TRAINING OF SCHOOL CHILDREN IN A MASSACHUSETTS TOWN BY EDWIN A. SHAW AND EDWARD A. LINCOLN PSYCHO-EDUCATIONAL CLINIC, HARVARD UNIVERSITY Series 1 No. 1 STUDIES IN EDUCATIONAL PSYCHOLOGY AND EDUCATIONAL MEASUREMENT Edited by WALTER F. DEARBORN MAY, 1922 Published by THE GRADUATE SCHOOL OF EDUCATION HARVARD UNIVERSITY, CAMBRIDGE, MASS. ^|fms?fi^ HARVARD MONOGRAPHS IN EDUCATION SERIES I Studies in Educational Psychology and Educational Measurement Mauuscripts for Series I should be addressed to Professor Walter F. Dear- born, Psycho-Educational Clinic, Palfrey House, Oxford Street, Cambridge 38, Mass. Remittances should be made by check or money order to The Graduate School of Education, HaTvard University, Cambridge 38, Mass. Series I of the Harvard Monographs in Education has been established for publishing the results of statistical and experimental studies and of educational tests in the general fields of educational psychology and educational measurement. The numbers are as follows: 1. A Comparison of the Intelligence and Training of School Children in a Massachusetts To-sra. E. A. Shaw and E. A. Lincoln. Postage prepaid, 50 cents. 2. The Marking System of the College Entrance Examination Board. L. Thomas Hopkins. Postage prepaid, 40 cents. 3. Stanaard Educational Tests in the Elementary Training Schools in Mis- souri. Walter P. Dearborn, Edward A. Lincoln and, Edwin A. Shaw. Postage prepaid, 60 cents. Whole No. 1 Series 1, No. 1 HARVARD MONOGRAPHS IN EDUCATION A COMPARISON OF THE INTELLIGENCE AND TRAINING OF SCHOOL CHILDREN IN A MASSACHUSETTS TOWN BY EDWIN A. SHAW AND EDWARD A. LINCOLN PSYCHO-EDUCATIONAL CLINIC, HARVARD UNIVERSITY Series 1 No. 1 STUDIES IN EDUCATIONAL PSYCHOLOGY AND EDUCATIONAL MEASUREMENT Edited by WALTER F. DEARBORN MAY, 1922 Published by THE GRADUATE SCHOOL OF EDUCATION HARVARD UNIVERSITY, CAMBRIDGE, MASS. COPYRIGHT 1922 By Wai/ter F. Dearborn )C1A692059 M 1 9 (323 'K6 j INTRODUCTION Most of the previously published School Surveys have dealt with the schools of a rather large community. This fact has undoubtedly led those interested in the schools of the smaller towns to feel that surveys were not practicable or valuable for any but the large systems with thousands of pupils, hundreds of teachers, a considerable invest- ment in school property, and plenty of money to meet the expenses of such investigations. This report is published in the hope that it will show the possibilities of the survey on a small scale. The school "sys- tem" with which the report deals consisted of only two schools, with scarcely more than 200 pupils, and less than 20 teachers. In spite of these small numbers and in spite of the fact that the survey was neces- sarily limited in its scope, it is the conviction of practically everyone concerned that the results have been of considerable value in improv- ing the efficiency of the schools. The cost was little more than a dollar a pupil. It is only fair to state that the community derived great advan- tages from the fact that the testing was done, papers were corrected, and the results were worked up as laborat&ry assignments of classes in Educational Measurement under the direction of the writers. This re- lieved the teachers of all work, and also materially cut down the finan- cial outlay necessary. However, it must be held in mind that the work could have been done by the regular teachers under the direction of the superintendent or some other leader who had given some time to the systematic study of Educational Measurement under a competent instructor. In such circumstances the cost would be even lower than it was in the present instance. Just as Part I of this report was ready for the press there came an invitation to repeat the survey. Some reorganization had taken place as the result of the findings which had been reported, and the School Board was desirous of learning the result of these changes. It was considered worth while to delay this report pending the prepara- tion of Part II, in order that the value of repeated measurements might be shown. The delay seems to have been justified, for Part II shows not only some changes which have come about as a result of the first survey, but also points out some new work which needs to be done in the schools. The two reports are printed exactly as they were sent to the School Board, with the exception that the class lists containing each iii pupil's record in the tests are omitted. A sample of them is shown in the appendix. The discussion, of course, has been made as non-tech- nical as possible, since the reports were intended primarily for the lay reader. The expert in educational measui-ement or the executive in a large school system will perhaps find little to interest him in this re- port. For those, however, who have felt that the paucity of pupils and the stringency of finances must deprive the small school system of the advantages of educational measurement we trust it may prove an inspiration and a guide. The authors wish to express their deep appreciation of the earnest and painstaking labor of their students in the preliminary work on which this report is based. They also acknowledge their indebtedness to Dr. Walter F. Dearborn for his helpful criticism in every phase of their undertaking. Edwin A. Shaw. Edward A. Lincoln. Psycho-Educational Clinic. Harvard Univeesity. IV REPORT ON THE RESULTS OF TESTING IN THE WAYLAND AND COCHITUATE SCHOOLS PART I— THE SURVEY OF 1921 PURPOSE OF THE REPORT It is the purpose of any school survey, whatever may be its nature and its scope, to discover and place at the disposal of all those vitally interested in any phase of the educational aeti\'ities of the community definite facts concerning the work which is going- on in the schools. To censure or to praise is not a primary function of the sui'vey ; its first and most important object must be to determine with absolute impartiality and scientific accuracy the existing conditions in the school or seliool system. Constructive criticism by disinterested out- siders is, of course, valuable and necessary, but teachers and adminis- trative officers, because they know in greater detail their schools and the community are, theoretically, at least, better able to make final decisions concerning the activities of the schools. This report, then, wiU place before the teachers, superintendent, school board, and citi- zens of the town certain facts about tlie pupils and the work of the schools with which they are so vitally concerned. Such comment and criticism will be made as, in the judgment of the surveyors, will con- tribute to the improvement of the educational opportunity offered at present by these schools. HOW THE SCHOOL WORK WAS MEASURED All modern school surveys make use of the standard tests in school subjects. As will be seen from the descriptions wliich appear later in the report, the standard tests are not in any great measure different from the oi'dinary school examinations which are more or less familiar to everyone who has ever attended school. The difference arises from the fact that the standard tests are scientifically constructed after long study and experimentation, and they have been given to many thous- ands of children all over the comitry, so that tlie average or standard performance in a room, grade, school, or whole system is pretty well known. Thus it is possible by the use of standard tests to compare the work of the various units of a school system, and also to compare the work done in the system with that done in any other where the same 2 A Comparison of the Intelligence and Training of tests have been used. For those reasons standard tests were used in the survey of the Wayland and Cochituate schools. WHAT TESTS WERE USED Although standard tests have been devised in a large number of the subjects commonly taught in the schools, not all of them could be used in this survey. Because of the limitations, tests were given in only the fundamental subjects; namely, Reading, Writing and Arithmetic. To these was added an intelligence test for the purpose of determining to some extent the character of the raw materials with which the various units of the school system have to work. Obviously, more is to be expected of a class composed of pupils of superior intelligence than from the average class, and inferior pupils cannot be expected to do as well as the average. HOW THE TESTS WERE GIVEN All the testing was done by examiners who had been well trained for the work at the Psycho-Educational Clinic, of the Harvard Grad- uate School of Education. Although the standard tests are for the most part so devised that they can be given by the average class-room teaclier, there is a technique of testing which is acquired only after considerable practice in the field, and it is, therefore, always desirable to have survey work done by trained examiners, as in the present in- stance. CORRECTION OF PAPERS The papers Avere all corrected by the clerks at the Psycho-Educa- tional Clinic. Here, also, much of the work can be done by the teacher, but it is more efficiently and economically done by trained experts. Moreover, when comparisons are to be made, it is an advantage to have the work done as far as possible by disinterested and impartial out- siders. PRESENTATION OP RESTILTS All the resiilts are presented in tabular form. The first set of tables presents the results of the tests by grade and school, and each table shows a distribution of all the marks obtained by the pupils of each grade, together with the grade median score. The Median rather than the Average is used because it is much more easily found, and be- cause most of the standard results are expressed as median scores. School Children in a Massachusetts Town 3 The median is simply the middle score when all the scores are arranged ill order of their size.* A second set of tables will show the Wayland and Cochituate me- dians together with those from other systems with which comparisons are desired, and in the Appendix will be found a sample of the class lists which will indicate the score of each pupil in each of the tests which he took. TABLE 1. DEARBORN GROUP INTELLIGENCE EXAMINATIONS General Examination I, May, 1921 Distribution of Scores by Grade t Cochituate Wayland Sub-Primary I II III I II III 120+ 4 no-119 2 100-109 1 2 3 90-99 2 2 1 2 80-89 3 1 1 70-79 1 1 4 2 1 4 60-69 1 3 4 1 1 Given 50-59 2 2 1 2 4 Series 40-49 4 3 1, 2 II 30-39 6 4 1 2 20-29 12 3 1 1 10-19 7 1 0-9 1 No. Cases 33 19 18 15 8 14 Median 28 48 75 105 43 70 'Thus the median of 5, 6, 8^ 11, 12, 15, 20 is 11. When there is an even number of items in the series the median is the same as it would be if one more item were added to the upper end of the series. Of the series 5, 6, 8, 10, 12, 14, 18, 19, 20, 21, the median is 14. fFor the lay reader who is not accustomed to reading tables of this sort it may be said that figures in the above table indicate the number of pupils who got each score on the examination. Thus in the Cochituate First grade 3 pupils made scores between 20 and 29, inclusive; 4 pupils made scores between 30 and 39, in- clusive; 3 scored between 40 and 49, inclusive, and so on. A Comparison of the Intelligence and Training of TABLE 2. DEARBORN GROUP INTELLIGENCE EXAMINATIONS General Examination I, May, 1921 Distribution of Scores by Age Cochituate Years 5 6 7 8 9 Score 120+ 1 4 1 110-1J9 2 100-109 1 3 90-99 1 2 2 80-89 1 1 2 70-79 2 1 3 60-69 2 2 3 2 50-59 3 1 1 40-49 1 2 2 2 30-39 5 2 3 20-29 8 4 2 2 10-19 2 3 1 0-9 No. Cases 12 20 15 20 13 Median 25 36 55 80 78 Wayland 6 7 1 1 1 1 2 2 2 2 2 1 2 1 1 1 1 5 10 6 53 50 75 RESULTS OP INTELLIGENCE EXAMINATIONS The Dearborn Group Intelligence Examinations which were used in this survey are made in two forms, one for the younger pupils who have not yet mastered the arts of reading and writing well enough to use them to any great practical extent, and the other for the older boys and girls. In the present instance the complete lower grade examina- tion was not used, but a shorter form, which has also been standardized, was employed. This examination will be referred to in this report as General Examination I. A distribution of the scores made on General Examination I in the different grades of both schools is presented in Table 1. BeeaiLse of the grouping of the grades it was found more practicable to test the tliird grade in Wayland with the Series II examination, so no results for this grade are to be found in Table 1. It will be seen that the pupils in the Cochituate school are, in both the first and second grades, somewhat superior to the pupils in the same grades at Wayland. We may reasonably expect, therefore, when we come to the subject-matter tests, to find the median scores of the first two grades in Cochituate somewhat higher than those of the corresponding grades in Wayland. School Children in a Massachusetts Town 5 In Table 2 the distribution of the General Examination I scores is according to the age of the pupils. The range in age is from 5 to 9 years in Cochituate, and from 6 to S only in Wayland. This difference is accounted for by the fact that in Cochituate there were Sub-Primary and Third grades tested, while at Wayland the test was given to the First and Second grades alone. TABLE 3. DEARBORN GROUP INTELLIGENCE EXAMINATIONS Series H, May, 1921 Distribution of Scores by Grade Cochituate Wayland Grades IV ' V V] [ VII VIII III IV V VI VII VIII H.S 200+ 190-199 1 180-189 2 170-179 1 5 160-169 2 3 150-1.59 1 9 140-149 1 1 3 130-139 1 2 1 3 120-129 1 4 1 4 1 5 110-119 1 1 4 2 1 1 5 100-109 1 2 2 1 4 3 2 90-99 2 2 1 2 1 1 1 3 1 2 80-89 2 1 1 3 2 1 70-79 7 3 1 2 1 2 1 2 1 60-69 2 1 1 1 3 1 1 50-59 7 . 2 2 1 1 1 1 40-49 5 ' 2 3 5 1 30-39 2 2 1 5 2 2 20-29 1 2 1 2 10-19 1 0-9 No. Cases 28 15 12 15 16 12 13 10 10 11 7 41 Median 50 72 : 100 ; 114 123 38 45 67 60 95 105 145 A Comparison of the Intelligence and Training of TABLE 4. DEARBORN GROUP INTELLIGENCE EXAMINATIONS Series n, May, 1921 Distribution of Scores by Age Cochituate Wayland Age 9 10 11 12 13 14 15 9 10 11 12 13 14 15 16 Score 200+ 190-199 1 2 180-189 170-179 1 1 3 160-169 1 2 1 150-159 1 1 1 140-149 1 1 1 2 130-139 1 3 2 1 120-129 1 4 3 1 1 2 110-139 2 3 1 1 1 1 2 1 100-109 1 2 1 1 1 1 2 1 1 90-99 2 1 1 2 1 1 2 2 1 80-89 2 1 2 1 1 1 70-79 3 2 3 1 1 2 o 1 1 60-69 2 3 1 50-59 5 3 2 2 1 1 40-49 5 2 1 1 5 3 1 80-39 1 2 2 2 1 1 20-29 2 1 1 1 1 10-19 1 0-9 No. Cases 16 12 10 12 13 15 6 13 12 5 11 10 11 8 9 Median 54 70 80 77 115 135 110 43 63 75 75 110 105 115 143 It will be seen that the sis-j^ear-old children in "Wayland make higher scores than the Cochituate jjupils of the same age, but that for the seven and eight-year-olds the Cochituate medians are better. It must he noted that there are so few six and eight year old pupils in Way- land that the medians for these years are rather unreliable.* However, the weight of the e^ddence points again to the fact that the Cochituate pupils are somewhat superior. Tables 3 and 4 show the distributions of the scores made on the Series II examinations which were given to the older pupils. Here again we find evidence of superiority of the Cochituate pupils, espe- *The reliability of a median is proportional to the number of oases from ■which it is obtained. See Thorndike, Mental and Social Measurements, Chap. XII. Eeliability of Measures. School Children in a Massachusetts Town 7 cially in the three upper grades. A noteworthy fact is the low standing of the sixth grade in "Wayland. The median for this grade is only 60, as compared with a median of 67 in the next lower grade. Other dis- crepancies of the same sort appear when the scores are distributed ac- cording to age in Table 4. The median score of the twelve year olds is no greater than that of the eleven year olds, and the thirteen year olds score, on the average, higher than the fourteen year olds. In the fif- teenth year the Wayland median is 5 points higher than that of Cochit- uate. This is explainable by the fact that in Cochituate the fifteen year olds are probably for the most part backward pupils who have not yet managed to get out of the grades, while the fifteen year olds from Way- land include the brighter pupils of that age who have gone on to High School. A Comparison of the Intelligence and Training of TABTiF! 5. PEET-DEARBORK ARITHMETIC TESTS Problems May, 1921 Cocliituate W ayland Grade IV V ' VI VII VIII III IV V VI VII VII Score 130 1 125-129 120-124 115-119 110-11-i 1 105-109 1 1 100-104 1 1 2 95-99 1 1 90-94 1 2 2 85-89 1 1 80-84 1 1 75-79 1 70-74 1 1 3 1 65-69 1 60-64 1 55-59 1 2 1 50-54 1 2 1 2 1 45-49 1 1 1 1 40-44 1 1 2 2 2 1 35-39 1 1 30-34 3 1 2 4 1 1 4 25-29 3 2 1 1 1 1 20-24 1 2 1 1 1 15-19 6 1 1 2 1 10-14 5 1 1 7 5 1 1 3 5-9 5 1 1 2 3 1 1 2 0-4 3 1 1 1 2 2 1 5 4 Total 28 13 12 15 16 12 17 10 10 13 10 Median 18 63 28 56 73 11 14 35 5 11 40 Standard 24 43 22 38 47 — 24 43 22 38 47 School Children in a Massachusetts Town 9 BESUIiTS OF THE ARITHMETIC TESTS For testing the accomplishment of the pupils in arithmetic the Peet- Dearborn tests were iised. These tests are a recent product which are a considerable improvement over the older standard arithmetic tests in that they are considerably more comprehensive in their scope. Previous tests in this field have dealt largely with the speed and accuracy of hand- ling whole numbers; the series used in these examinations includes all the leading problems which a pupil needs to master, such, for example, as the use of the zero, placing decimal points, the use of round num- bers, and the use of common fractious. The problems in each test range from very simple ones to those which require superior intelligence for solution, and in order to get a just method of scoring, each problem is given a value in proportion to its difficulty, which was statistically de- termined. The tests run in two series; one for grades Four and Five, and the other for grades Six, Seven and Eight. The distributions of the arithmetic scores and the median scores of the various grades are shown in Tables 5 to 9 inclusive. It will be noted in all these tables that the Sixth grade scores are lower than those of the Fifth grade. This is because of the different series employed in the examination. In the Arithmetic Problems (Table 5) the Cochituate scores in each series are considerablj' higher than the Wayland scores. This dif- ference amounts to something more than a full grade,that is, the Sixth grade at Wajdand does not do so well, on the average, as the Fifth grade in Cochituate. The Wayland Sixth grade did especially poor work, as half the class failed to solve correctly more than one problem. 10 A Comparison of the Intelligence and Training of TABLE 1 6: PEET-DEABBOBN ARITHMETIC TESTS Addition May, 1921 Cochituate Wa; yland Grade IV V VI VII VIII : [II IV V VI VII VII Score 130 125-129 120-124 115-119 1 1 110-114 1 1 105-109 1 100-104 1 1 2 1 95-99 1 90-94 1 85-89 1 1 80-84 1 75-79 1 2 70-74 1 1 1 65-69 3 1 1 60-64 1 4 55-59 1 1 2 1 2 3 50-54 1 1 2 1 45-49 2 1 3 2 40-44 3 2 3 1 2 35-39 4 2 1 1 2 30-34 2 4 2 4 1 3 25-29 2 1 1 4 1 20-24 3 1 1 1 1 1 1 2 15-19 6 2 2 1 1 1 1 10-14 2 1 1 5-9 2 6 3 1 0-4 1 1 2 1 2 Total 28 13 12 15 16 12 17 10 10 13 10 Median 30 43 38 63 60 9 26 45 40 28 58 Standard 30 51 35 51 54 30 51 34 51 54 School Children in a Massachusetts Town 11 TABLE 7. : PEET -DEAJ XBORJ i ARl :THj\n 5TIC ' TESTS 5 Subtraction May, 1921 Cochituate Wayland Grade IV V VI VII VIII III IV V VI VII VII] Score 130 1 1 1 125-129 120-124 1 115-119 1 1 1 110-lM 105-109 1 1 100-104 " 1 95-99 1 1 3 90-94 1 1 85-89 1 1 2 2 1 1 80-84 1 75-79 1 1 1 1 1 3 70-74 1 1 1 1 65-69 1 1 2 1 1 60-64 2 2 1 3 1 1 55-59 1 1 1 50-54 2 1 1 1 2 1 45-49 1 1 1 2 40-44 5 3 2 1 35-39 4 2 1 30-34 2 5 4 4 1 1 1 25-29 1 3 1 20-24 2 1 4 1 1 1 15-19 3 1 1 2 2 2 2 10-14 1 1 2 2 5-9 1 4 2 1 0-4 4 1 2 1 Total 28 13 12 15 16 12 17 10 10 13 10 Median 35 64 40 78 80 13 32 55 15 38 55 Standard 29 51 34 52 62 29 51 34 52 62 12 A Comparison of the Intelligence and Training of TABLE 8. PEET-DEABBORN ABITHIVIETIC TESTS Multiplication May, 1921 Cochituate Wayland Grade IV V VI VII VIII III IV V VI VII VII Score 130 1 2 1 125-129 120-124 1 115-119 2 1 110-114 1 105-109 1 100-104 2 95-99 1 1 3 1 90-94 1 3 1 2 1 85-89 1 1 1 2 2 80-84 1 1 75-79 1 1 1 70-74 1 1 65-69 1 1 60-64 1 4 1 1 1 1 55-59 1 1 2 1 2 50-54 1 1 1 2 1 1 45-59 1 1 2 1 1 40-44 1 1 4 1 1 1 35-39 1 1 1 1 2 30-34 2 1 3 4 4 25-29 1 2 1 1 1 20-24 4 1 2 2 2 1 15-19 3 5 2 1 10-14 3 2 1 5-9 2 6 0-4 7 1 1 1 3 Total 28 13 12 15 16 12 17 10 10 13 10 Median 18 91 40 63 97 9 33 75 31 43 55 Standard 28 48 41 52 58 28 48 41 52 58 School Children in a Massachusetts Town 13 TABLE 9. PEET-DEARBORN ARITHMETIC TESTS Division May, 1921 Cochituate Wa; yland Grade IV V VI VII VIII III IV V VI VII VIII Score 130 5 125-129 120-124 115-119 2 1 110-114 2 105-109 1 100-104 95-99 1 90-94 85-89 2 1 2 80-84 75-79 1 1 4 1 2 1 70-74 3 3 1 2 2 65-69 1 1 1 60-64 1 1 2 3 1 1 2 55-59 1 1 50-54 2 1 1 1 1 1 45-49 2 1 1 1 2 40-44 5 1 1 1 3 2 35-39 2 1 1 1 30-34 5 1 1 1 1 25-29 1 1 3 1 20-24 3 1 3 4 1 3 1 15-19 3 1 2 6 2 10-14 3 1 1 2 5-9 1 5 3 2 1 0-4 1 1 1 2 1 1 Total 28 13 12 15 16 12 17 10 11 13 10 Median 32 71 63 76 75 10 18 73 18 38 53 Standard 28 49 31 49 61 28 49 31 49 61 The facts concerning the Addition examples are shown in Table 6. Here the discrepancies between the corresponding grades of the two schools are not great, except in the Seventh grade scores. In this grade the Cochituate median is 35 points higher than the Wayland median. Table 7 shows the results of the Subtraction test. Here again there are variations in favor of Cochituate which range from 6 to 27 points. 14 A Comparison of the Intelligence and Training of In tlie Multiplication test (Table 8) the Fourth grade in Cochituate did very much poorer work than it did on the previous tests, and the "Waylaud median is higher in this grade. In all the other grades, how- ever, Cochituate did considerably superior work, with medians ranging from 9 to 27 points higher than those of the Waylaud classes. The test in Di%asion (Table 9) shows the Cochituate work superior except in the Fifth grade. It is also worthy of note that the Eighth grade in Cochituate does not attain a higher median than the Seventh grade. It is, however, considerably above the standard median. In summary of the results of the Arithmetic tests it may be said that the work in Cochituate is considerably in advance of that at Way- land, with only a few exceptions. The difference amounts, on the aver- age, to about a year's work. This is the state of affairs which was pre- dicted upon examination of the results of the intelligence tests, and it is not, therefore, surprising. TABLE 10. PICTURE SUPPLEMENT SILENT READING TEST Distribution of Scores May, 1921 Cochituate Wayland Grade III IV V VI VII VIII III IV V VI VII VIII 100 98 92 86 1 80 1 74 1 1 68 4 1 1 62 1 3 1 1 56 2 4 1 1 2 1 50 3 5 2 2 1 1 1 44 1 3 1 1 3 1 2 1 2 2 38 2 1 1 2 2 3 2 2 32 4 1 1 4 1 4 1 1 26 3 1 3 5 4 5 2 1 1 1 20 1 1 2 3 1 1 2 5 14 1 1 3 8 1 1 1 1 2 1 1 1 2 2 No. Cases 19 23 14 10 13 17 11 16 10 8 13 9 Median 38 56 20 25 26 32 26 32 38 20 20 38 School Children in a Massachusetts Town 15 RESUliTS OP THE READING TEST The test used in this survey for the measurement of reading is one recently published by the Russell Sage Foundation, and is knowu as Picture Supplement Scale, Form 1. It consists, according to the au- thor's description, of "a series of pictures and paragraphs about them. These paragraphs consist of instructions which the pupil follows by marking with his pencil a line or lines to supplement the picture. His ability to do this in accordance with printed insti-uctions reflects the rapidity and accuracy with which he can read." A pupil's score is the number of paragraphs he marks correctly, but in order that all the grades may be put on the same basis the author provided a table of credits corresponding to the number of paragraphs correctly marked in each grade. Tlie distribution of the scores on the credit basis is given in Table 10, together with the median credit attained in each grade. For the most part, the Coehituate median credits are the higher, but in the Eighth grade Wayland has the slight advantage of one more paragraph correctly marked. This result again accords with the findings of the intelligence examinations which led us to expect better work from the pupils in Coehituate. RESULTS OF THE PENMANSHIP TESTS The Holmes Test for the Speed and Quality of writing was used in this investigation. For the speed test, a short sentence, made up of simple and familiar words, but involving nearly all the letters and a large number of the common letter combinations, is written repeatedly, first for a one minute pei'iod, and then for four minutes. The sentence is practically learned by the children before they begin to write on the test, .so that little time is lost because of forgotten copy. The speed of writing is obtained by taking the average of the number of letters writ- ten in the one-minute period and the number written per minute in the four minute period. 16 A Comparison of the Intelligence and Training of TABLE 11. ho: LMEi 5 TEl ST, SI >EEI Coehitiiate Grade Ill IV V VI VII VI] Letters per mill. 120+ 5 115-119 2 1 110-114 3 4 105-109 1 1 2 100-104 1 1 1 2 95-99 2 2 1 90-94 1 1 1 85-89 1 2 1 2 80-84 1 2 1 75-79 1 3 70-74 2 5 3 65-69 5 2 1 60-64 2 3 1 55-59 5 1 50-54 2 1 45-49 1 5 1 1 40-44 4 1 35-39 30-34 1 25-29 1 20-24 19- No. Cases 15 27 14 10 14 16 Median 60 61 77 85 100 113 Wayland V VI VII VIII 1 1 1 1 1 1 1 1 1 2 3 1 2 1 1 1 2 1 1 1 3 1 4 2 2 2 2 2 1 3 3 1 3 1 1 2 1 1 1 2 1 1 11 15 10 9 48 53 68 63 11 9 98 School Children in a Massachusetts Town TABLE 12. HOLMES TEST, QUALITY OF WRITING. May, 1921 Cocliituate Wayland 17 Grade II III IV V VI VII vii: [III IV V VI VII VII 90-94 85-89 80-84 1 75-79 1 1 70-74 2 3 1 1 65-69 2 60-64 3 1 2 1 1 1 55-59 2 1 2 1 1 3 50-54 3 2 1 1 3 7 1 3 2 1 45-49 2 8 5 3 3 3 2 4 2 2 40-44 3 4 9 3 4 2 1 2 5 2 35-39 1 4 1 3 1 1 3 3 1 1 30-34 6 2 1 2 4 4 1 25-29 4 4 2 1 20-24 2 15-19 10-14 Total 16 19 23 14 10 14 16 11 16 10 8 10 9 Median 32 39 46 43 45 50 53 32 40 60 48 50 57 The facts concerning the speed of writing are shown in Table 11. Here again the medians show that the work at Wayland is about a grade below that at Cochituate. The low standing of the Sixth grade at Way- land is once more noticeable. To get the quality of penmanship by the Holmes test, two samples of each child's writing are graded on the Ayres Handwriting scale. The first of these samples is the last half of the material wi-itten in the four minute speed test, and for the other the pupils are given a short passage to write from dictation. The samples are graded separately, each by two correctors working independently, and the average of the four marks (two on each paper) is taken as the quality of the pupil's handwriting. The distributions and medians of the marks obtained in the quality of penmanship are shown in Table 12. In this test we find for the first time that the Wayland medians equal or surpass those of Cochituate. The difference in most cases is small, amounting to about half a step on the Ayres Scale, but in the Fifth grade the penmanship at Wayland is especially good. Grade IV Cochituate 18 Wayland 14 Standard 24 VI VII VIII 28 56 73 5 11 40 22 38 47 18 A Comparison of the Intelligence and Training of COMPARISONS WITH OTHER SYSTEMS 1. The Arithmetic Tests In the following tables, 13 to 17 iiaclusive, are shown the Cochituate and Wayland median Arithmetic scores compared with the standards. TABLE IS. PROBLEMS May, 1921 V 63 35 43 In the Fourth grade neither school attains the standard median score for the Problem Test, but in the Fifth grade the Cochituate me- dian is considerably above the standard, while the Wayland median is considerably lower. In the upper grade test which was given in the grades above the Fifth, the Cochituate medians are uniformly consid- erably larger than the standards, and the Wayland medians are consid- erably smaller. TABLE 14. ADDITION TEST May, 1921 V 43 45 51 In the Fourth grade, the Cochituate median in Addition exactly equals the standard, but the Wayland score falls below. Neither school equals the Fiftli grade standard. In the Sixth and Seventh grades the Cochituate medians surpass the standards, while the Wayland scores fall below again, especially in the Seventh grade. In the Eighth grade, both scores exceed the standard by a good margin. Grade IV Cochituate 30 Wayland 26 Standard 30 VI VII VIII 38 G3 60 30 28 58 35 51 54 TABLE 15. SUBTRACTION TEST May, 1921 Grade IV V VI VII VIII Cochituate 35 64 40 78 80 Wayland 32 55 15 88 55 Standard 29 51 34 52 62 In grades Four and Five the Subtraction medians in both schools are higher than the standards. In the upper grade test we find the Cochituate medians exceeding the standards by a good margin, espe- cially in the Seventh and Eighth grades. Wayland in the upper grades falls below the standard, and has an especially low score in grade Six. School Children in a Massachusetts Town 19 TABLE 16. MIILTIPLICATION TEST May, 1921 V 91 75 48 In multiplication the Cochituate Grade Four median falls below the standard, but Wayland is above it. In the Fifth grade both schools exceed the standard by large variations. The upper Wayland grades cannot attain the standards in this operation, but the Eighth grade comes very close to it. The Cochituate median is practically the same as the standard in the Sixth grade, and in the Seventh and Eighth grades the medians of this school are especially high. Grade IV Cochituate 18 Wayland 33 Standard 28 VI VII VIII 40 63 97 31 43 55 41 52 58 TABT,E 17. DIVISION May, 1921 TEST Grade IV V VI VII VIII Cochituate 32 71 63 76 75 Wayland 18 73 18 38 53 Standard 38 49 31 49 61 In the Fourth grade neither school attains the Division standard, though the Cochituate median is very close. Both schools are weU above the standard in the Fifth grade. In the three upper grades we again find that the Cochituate medians are well above the standards for those grades, while the Wayland medians fall below. 20 A Comparison of the Intelligence and Training of TABLE 18. PICTURE SUPPLEMENT, SILENT READING TEST Form I, May, 1921 Per Cent of Pupils Receiving Each Mark Cochituate Wa ylauc I Grade in IV V VI VII VIII Standard III IV V VI VII VIll 100 98 1— 92 1 + 86 3 80 5 4 74 5 4 10 6 68 18 6 8 62 5 13 10 8 10 10 56 11 18 7 11 6 20 8 50 16 22 7 20 15 6 12 9 44 5 13 10 8 18 11 9 13 10 15 11 38 11 4 7 6 10 18 13 30 25 22 32 21 4 7 8 23 8 9 25 10 22 26 16 4 15 30 39 23 6 46 13 13 8 11 20 5 50 10 15 18 4 6 10 25 38 14 7 8 3 6 23 8 1+ 6 10 13 11 2 10 1— 3 9 13 25 No. Cases 19 23 14 10 13 17 11 16 10 8 13 2. The Reading Test Table 18 sets forth a comparison of the scores in the Silent Reading Test with the standards. The scoring of this test is so arranged that in any average grade from the Second to the Eighth inclusive the percent- age of pupils obtaining each score is the same. It will be readily seen by reference to the Table that in neither school does the distribution equal or approximate the standai'd. There is no grade in either school which has the average quota in the upper third of the distribution. 3. The Penmanship Tests In Table 19 are shown the medians for the speed of wi'iting. To- gether with the standards there are in this Table the scores from three Massachusetts communities where the Holmes Test has been given. School Children in a Massachusetts Town 21 TABLE 19. SPEED OF PENMANSHIP May, 1921 Grade III IV Y VI VII VIII Coehituate 60 61 77 85 100 113 Wayland 48 53 68 63 88 98 Standard 49 62 76 87 90 98 Newton 55 59 73 85 94 102 Brookline 76 87 90 98 Fall River 86 89 The Coehituate medians surpass the standards by a good margin in every grade except the Fifth, and also exceed the medians of the other school systems which are listed in the Table. Wayland, on the otlier band, shows metlian scores considerably below the standards of the other school systems up until the Seventh and Eighth grades, where the stand- ards are equalled. The medians for the quality of penmanship are shown in Table 20. Here we find that, in general, the medians are higher than the standards, and higher than those obtained in Brookline and Fall River. They do not quite equal the Newton scores in the grades below the Seventh XABTE SO. QUALITY OF PENMANSHIP May, 1921 Grade Ill IV V VI VII VIII Coehituate 39 46 43 45 50 53 "Wayland 32 40 60 48 50 57 Standard 36 39 44 46 47 49 Newton 50 45 48 51 50 53 Brookline 44 46 47 49 Fall River 44 47 SUMMARY Our comparison of the two schools shows, in general, with the excep- tion of a grade or two here and there, that the quality of work done by the Coehituate pupils is decidedly superior to that done at Wayland. This statement holds except for the quality of penmanship, in which Wayland does especially well. An explanation of the better perform- ances of the Coehituate pupils is found iu the fact that they make higher scores on the Intelligence Examinations, thus indicating the probability that they are somewhat superior in their native abilities. It is possible also to make a rather general statement about the com- parisons of the Coehituate and Wayland scores with the standards. For the most part, the Coehituate medians are equal to, or better than the 22 A Comparison of the Intelligence and Training of standards, while the Waylaud scores equal the standards in only a few cases out of the total. Again there is an exception to this in the case of the quality of penmanship, in which test the Wayland pupils equalled or surpassed the standards in most of the grades. SUGGESTIONS It is a fact sometimes overlooked in a report of this sort that inter- pretation of the results set forth is chiefly the business of the teachers and supervisory officials of the school system in which the tests have been given. No one else is continually on the ground and familiar with all the conditions ; no one else can understand the facts in their full sig- nificance. Perhaps the most valuable use to which standard tests may be put by a school system is that of serving as a guide for the formulation of standards for its own work. Tests are first given to show the actual attainment of the pupils in the schools. With these results and informa- tion concerning the results of testing in other schools standards may be set in the various school subjects. It must be held in mind continually, when such standards are being determined, that they must always be made with reference to the condi- tions in the system in question. Each community has its own individ- uality — its own problems, its own potentialities, and its own limitations. The recent development of intelligence tests has demonstrated that there are undoubtedly differences in native capacity which appear to exist between schools and school systems as well as between individuals. No system, therefore, may safely take as its own, standards which have been made elsewhere, but each must work out a set of standards for itself. In this connection it should be pointed out that there has been some- what of a tendency to set too high the standards in the school subject tests. It is forgotten sometimes that these tests, for the most part, measure only the simpler and more mechanical phases of a subject, and that excessive attention to these phases may mean that more important things are neglected. Thus it becomes a question, for instance, whether the relatively high scores made by the Cochituate pupils on most of the tests show a desirable state of affairs, and whether the school should aim to keep future classes at the same level. If the superior scores are due to superior ability, as may l)e the case, it is to be desired that the same liigh grade of work be continued, but if these scores are due to over-em- phasis on the mechanical parts of the various subjects, then some change should be made. This brings us to another important consideration : namely, that the standards should always be tentative. It is impossible to determine once and for all what the accomplishments of pupils should be. From School Children in a Massachusetts Town 23 time to time, as the community clianges, the standards should change also. In other words, the problem of standards is always an open and live one. Once tentative standards have been set, the tests are useful to deter- mine the progress of the various schools in relation to those standards, and to gather material for the revision of them. For these pm-poses the tests should be repeated at not too frequent intervals; once a year is probably often enough. The results should be studied with the purpose of picking out schools, classes and individual pupils to which some sort of special atten- tion should be given. One school of a system may do especially well in writing, but fall down in reading, while another may exceed the stand- ards in both reading and writing, but make a very poor showing in the arithmetic. A class may stand high in addition, but fail in division. Some pupils will be found to be in need of much individual work, while others will seem to have attained satisfactory proficiency in all lines. In the lower part of the gi'oup thei'e wiU be found cases which should be turned over to a specialist for searching mental examinations, while there will be others whose attainments will warrant double promotion, or who should be enabled to speed up their work in some other way. Superintendents and teachers should remember that neither every individual nor every class should be expected to attain the standards. If anything stands out clearly in testing, it is the fact of individual differ- ences, and many pupils will never be able to reach the standard medians. Indeed, the very definition of the term "median" indicates that it is a mark which only one-half of the pupils equal or exceed. It is often the practice of supervising officials to judge the worth of their teachers by the scores that the pupils in the various cla.sses make on the standard tests. It is more likely, however, that the success which a class has in making high scores in the standard tests has little or nothing to do with the ability of the teacher. Since the assignment of teachers to classes is largely a matter of chance, it may easily happen that the best class in the school may have the poorest teacher, and that the success of a class may be in spite of the teaching, rather than because of it. To sum up, then, if the facts set forth in this report are to become useful they must be studied carefully and exhaustively by the people in- terested in the schools. It must be remembered that high scores in the tests are not the only, nor even the most desirable ends of school work, and that possibly the classes which do poorly in these are getting fully as 24 A Comparison of the Intelligence and Training of much from the school. The results in the subject tests must be con- sidered in relation to the intelligence scores, and to all other facts con- cerning the general conditions in the schools and in the community. A little follow-up work, -with special attention along the lines suggested above can hardly fail to be of inestimable value to the pupil, the teacher, and the society of which the pupil is to become a part. % School Children in a Massachusetts Town 25 PART II— THE SURVEY OF 1922 In May, 1922, a year after the first tests had been given in "Wayland and Coehitnate the whole survey was repeated. The tests used were the f^ame, except that tliere had been some slight modifications of the Intelli- gence and Arithmetic Tests. The testing and grading was again done by carefully trained examiners. The results of this second testing are set forth in tables which are numbered to correspond with the tables in Part I, so that it will be easy for anyone who is interested to compare the distributions for the two years. RESULTS OF INTELLIGENCE EXA»nNATIONS Distributions of the scores made by the pupils of the lower grades on the Dearborn General Examination A are shown in Table la. In Table 2a the same scores are distributed according to the ages of the pupils. There is evidence in these figures that Cochituate has pupils of su- perior mentality in the First grade, as shown by the median score of 48 as opposed to the Wayland median of 33. Also, it may be seen in the second Table that the median score of the six-year-olds is identical in both schools. In the Second grade, however, it appears that the Way- laud children are superior, with a class median of 66 against 60 made by the Cochituate class. The seven-year-olds at Wayland also do somewhat better than the seven-yeai*-olds at Cochituate. The superiority of the Wayland children in the Second grade amounts to about five months on a mental age basis, and while this is not an especially great difference, it is probably large enough to be significant. 26 A Comparison of the Intelligence and Training of TABLE la. DEARBORN GROUP INTELLIGENCE EXAMINATIONS General Examination A, May, 1922 Distribution of Scores by Grade Cochituate "Wayland Snb-Primary I II III I II III 95-99 2 90-94 1 3 85-89 3 80-84 1 4 75-79 3 3 70-74 2 1 65-69 2 2 2 60-64 2 3 1 Given 55-59 1 6 3 2 2 Series 50-54 3 1 II 45-49 4 1 1 1 1 40-44 2 6 1 35-39 2 2 4 30-34 6 2 3 25-29 2 1 20-24 1 1 15-19 10-14 1 5-9 0-4 1 No. Cases 22 29 11 24 12 9 Median 30 48 60 80 33 66 Mental Age of Median 6-8 7-7 8-4 10-5 6-10 8-9 Child School Children in a Massachusetts Toivn 27 TABLE 2a. DEARBORN GROIT IXTELLIGENCE EXAMINATIONS General Exanuuation A, May, 1922 Dlsti •ibution of Scores by Age Cocliituate Wayland Years 5 6 7 8 9 10 6 7 8 95-99 1 90-94 1 3 1 85-89 2 1 80-84 1 2 75 79 2 1 2 1 70-74 1 1 1 6569 1 2 1 1 1 60-64 1 1 2 1 55-59 4 2 3 2 1 1 1 50-54 3 1 45-49 1 1 3 1 1 1 40-44 1 1 4 3 35-39 1 3 1 3 30-34 2 3 2 1 2 1 25-29 1 7 1 1 20-24 2 1 15-19 1 10-14 1 1 5-9 1 0-4 1 No. Cases 6 20 21 18 11 7 6 10 5 Median 30 30 48 56 73 78 30 55 58 M.A. of Med. Child 6-8 6-8 7-7 8-1 9-6 10-1 6-8 8-0 8-2 NOTE : — Two 11-year old pupils and one 13-year old from the Coehit- uate school are omitted from this Table. 28 A Comparison of the Intelligence and Training of In Tables 3a and 4a will be found the distributions of the Dearborn General Examination C which was given to the older children. In the Fourth and Sixth grades the Wayland medians are higher, in the Fifth and Seventh grades the Cochituate medians are higher, while in the Eighth grade the medians are exactly the same. The largest difference is in the Sixth grade where Wayland has a median which is 8 points higher than that of Cochituate. This means a difference of over a year in terms of mental age, and is a large variation. The difference of four points in the Fourth grade is probably also large enough to be of signifi- cance. This state of affairs is quite different from that which was discov- ered the previous year. In 1921 the results showed rather clearly that in most grades the pupils of the Cochituate school were superior to those in the same grade at Wayland. Now many of the differences seem to have disappeared, and those which remain seem to be in favor of Way- land rather than of Cochituate. A possible explanation of this change lies in the fact that the results of the first report were used to make ex- tensive changes in tlie organization of each school, and it seems likely that the reorganization classified the children into much more homoge- neous groups. One interesting fact should be noted in passing. In 1921 the Sixth grade at Wayland proved to be inferior to the fifth grade in the intelli- gence test. In 1922 the Seventh grade, composed largely of the same pupils who made up the Sixth grade in the previous year, is quite mark- edly inferior to the Sixth. Evidently the character of this class was not entirely changed in the reorganization. Sciiool Children in a Massachusetts Town 29 TABLE 3a. DEARBORN GROUP INTELLIGENCE EXAMINATIONS Gen. Exam. C, May, 1932 Distribution of Scores by Grade Coeliituate Wayland III IV V VI VII VIII Grade IV V VI VII VII 80+ 75-79 70-74 1 65-69 1 60-64 1 2 3 55-59 1 1 2 50-54 1 1 1 4 45-49 2 2 40-44 3 4 4 2 35-39 1 2 3 1 1 30-34 3 6 2 3 1 25-29 2 6 1 20-24 2 6 1 1 15-19 3 8 2 10-14 2 1 1 5-9 2 0-4 2 No. Cases 17 35 17 13 17 Median 19 27 38 42 53 M.A. of Med. Child 10-2 11-6 13-4 14-0 15-1^ 2 1 1 1 1 5 1 3 2 2 o 3 1 2 1 1 1 1 2 1 3 1 2 3 2 1 2 3 4 1 6 2 1 1 2 2 1 1 1 1 13 11 12 11 12 15 18 23 25 46 40 53 10-0 10-10 11-2 14-8 13-8 15-10 30 A Comparison of the Intelligence and Training of TABLE 4a. DEARBORN GROUP INTELLIGENCE EXAAUNATIONS Gen. Exam. C, May, 1932 Distribution of Scores by Age Cochituate Wayland Age 9 10 11 12 13 14 15 8 9 10 11 12 13 14 15 80+ 75-79 70-74 1 65-69 1 60-64 1 2 2 1 1 2 55-59 1 1 2 1 1 1 1 3 50-54 1 2 1 3 1 1 2 45-49 2 1 1 2 1 1 3 2 40-44 2 1 5 2 3 2 2 35-39 1 2 2 3 1 1 30-34 2 4 1 3 1 4 1 1 2 1 1 1 1 25-29 1 3 3 1 1 4 2 2 20-24 2 2 3 1 2 2 2 4 1 15-19 3 4 3 3 6 2 1 2 1 10-14 2 1 1 1 1 1 1 5-9 2 1 1 0-4 1 1 No. Cases 9 23 16 20 14 16 1 9 9 15 10 8 10 6 5 Med'n 24 26 25 40 40 45 18 26 24 30 45 43 50 M.A. of Med. Child 11-0 11-4 11-2 13-8 13-8 14-6 10-0 11-4 11-0 12-0 14-6 14-2 15-4 NOTE : Oue 16-year old pupil and one 18-year old pupil from the Way- land school have been omitted from this Table. School Children in a Massachusetts Town 31 RESITLTS OP THE ABITHJLETIC TESTS As previously noted, the Peet-Dearborn Arithmetic Tests had been somewhat modified between the two testings so the results of the two years are not directly comparable. There are still, however, two series of the tests, one for the Fourth and Fifth grades, and the other for grades Six, Seven and Eight. This accounts for the fact that in all the Tables the Sixth grade scores are usually lower than those made by the Fifth grade. In Table 5a wiU be found the distributions of the scores obtained on the Problem Test, together with the class medians. The medians in grades IV and V show better work by the Wayland children, though the Fifth grade medians differ by only two points, an insignificant differ- ence. In the upper grades, however, the scores are very decidedly in favor of Cochituate. These differences cannot be explained on the basis of the intelligence of the classes, except perhaps in the Fourth grade where the Wayland class made a significantly higher median in the intelligence test. It would seem that in the Sixth grade especially, where the mental age of the Wayland pupils is about a year higher, a larger median might reasonably be expected. Table 6a presents the distributions and medians in the Addition Test. Here we find that the medians for the Fourth and Fifth grades are practically the same, though in the former grade the Cochituate median is somewhat higher. In the upper grades we find the Cochit- uate medians are larger bj' substantial margins which range from 10 to 21 points. In Table 7a we have the facts about the Subtraction Test. It wiU be seen that in the three lower grades the Wajdand medians are decid- edly higher, while in the Seventh and Eighth grades the Cochituate classes show to much better advantage. It appears that the Seventh and Eighth grade medians at Wayland are extremely low. 32 A Comparison of the Intelligence and Training of TAB1,E 5a. PEJ ET-DI 3ARB( )RN r »ROGE Problems May, 1922 Cochituate Grades IV V VI VII VIII 130+ 2 125-129 120-124 115-119 110-114 1 105-109 100-104 95-99 2 90-94 5 85-89 1 80-84 2 1 1 75-79 1 1 2 70-74 4 1 1 65-69 2 1 60-64 2 55-59 1 3 3 1 50-54 1 4 4 1 45-49 1 40-44 5 1 35-39 2 30-34 25-29 1 2 2 1 20-24 3 5 2 1 1 15-19 2 2 10-14 3 2 1 5-9 5 2 0-4 2 3 No. Cases 17 33 17 13 17 Median 13 43 52 52 91 Standard 24 38 22 41 56 Wayland IV V VI VII VIII 1 3 1 2 1 2 10 12 2 10 2 2 2 3 110 12 10 10 11 2 10 10 11 10 2 2 1 4 12 1 1 11 12 11 12 15 22 50 33 40 44 24 38 22 41 56 School Children in a Massachusetts Town 33 TABLE 6a. PEET-DEARBORN PROGRESS TESTS IN ARITHMETIC Addition May, 1923 Coehituate Wayland Grades IV V VI VI VIII IV V VI VII VII 130+ 1 125-129 120-124 2 115-119 1 1 110-114 105-109 100-104 2 95-99 1 1 90-94 85-89 1 4 1 4 2 80-84 1 75-79 2 1 2 2 70-74 1 1 1 1 2 2 1 1 65-69 2 2 1 1 60-64 1 5 2 5 3 1 4 2 55-59 1 2 1 2 1 50-54 2 2 1 1 45-49 2 2 1 2 1 40-44 3 1 5 1 2 35-39 1 2 1 2 1 1 2 30-34 2 6 1 2 3 1 1 1 25-29 2 1 20-24 5 2 2 1 4 3 15-19 1 3 1 2 1 10-14 1 5-9 0-4 1 N.o Cases 17 33 17 13 17 11 12 11 12 15 Median 31 62 49 54 69 24 64 39 38 48 Standard 30 50 35 52 60 30 50 35 52 60 A Comparison of the Intelligence and Training of TABLE 7a. Grades 130+ 125-129 120-124: 115-119 110-114 105-109 100-104 95-99 90-94 85-89 80-84 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0-4 No. Cases Median Standards PEET-DEARBORN PROGRESS TESTS IN ARITHMETIC Subtraction May, 1922 Cochituate Wayland IV V VI VII VIII IV V VI VII VIII 1 1 1 2 3 1 1 1 5 1 17 38 28 3 1 1 4 3 4 3 2 2 1 3 1 1 1 32 58 47 2 3 3 1 2 1 1 1 1 16 52 39 1 2 3 1 2 1 1 1 1 13 53 57 1 1 3 1 1 2 2 3 1 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 17 68 68 11 51 28 12 65 47 1 3 1 11 63 39 2 1 3 2 1 1 1 1 12 25 57 1 1 1 1 1 1 1 3 2 3 15 34 66 School Children in a Massachusetts Town 35 Distributions and medians for the Multiplication Test are shown in Table 8a. The Fourth grade medians are practically identical. Way- land has an advantage of 8 points in the Fifth grade, and Cochituate excels in the upper grades by margins running up to 32 points in grade Eight. Table 9a shows the results of the Division Test. Here we find that the Cochituate medians are decidedly higher throughout the whole five grades. The points of difference range from 9 in the Sixth grade to 38 in the Eighth. In Table 9b the averages of the pupils in all operations are distrib- uted. The medians are substantially the same in the Fourth and Fifth grades, but in the upper grades the Cochituate medians are considerably higher. Summarizing the results of the Arithmetic Tests we may say, in general, that the work of the Cochituate childi'en is superior. This is not true in all cases in the two lower grades (IV and V) where the me- dians are more nearly equal and where the Wayland medians are in one or two cases distinctly higher. The differences which we find are not to be explained on the basis of intelligence differences, as the latter are small except in cases where they favor Wayland. It seems likely, there- fore, either that more time is given to arithmetic, especially in the upper grades, at Cochituate than at Wayland, or that the Cochituate teachers use markedly superior methods in their teaching. 36 A Comparison of the Intelligence and Training of TABLE 8a. PEl DT-DE ABB( )RN P ROGRESl S TES' TS IN ABII rHME TIC Multiplication May, 1922 Cochituate Wayland Grades IV V VI VII VIII IV V VI VII VIII 130+ 1 125-129 120-124 115-119 110-114 1 1 105-109 2 100-104 3 3 2 1 95-99 1 1 90-94 1 1 85-89 6 3 3 2 1 1 80-84 1 1 1 1 75-79 3 3 2 1 1 70-74 1 1 2 3 2 65-69 1 1 60-64 2 5 1 2 1 55-59 2 2 1 50-54 3 1 1 1 1 1 2 45-49 2 3 1 1 1 40-44 2 2 1 4 1 1 2 3 35-39 3 1 2 1 30-34 2 3 4 2 1 2 2 2 25-29 1 1 1 1 1 2 20-24 1 1 1 2 2 15-19 1 1 1 1 10-14 1 1 5-9 0-4 1 No. Cases 17 32 16 13 17 11 11 11 12 15 Median 44 75 60 76 83 42 83 39 30 51 Standard 28 48 31 51 68 28 48 31 51 68 School Children in a Massachusetts Town 37 TABLE 9a. PEI 5T-DE ARBO iRN P ROGRES IS TESTS IN ARITHlVrETIC Division May, 1922 Cochituate Wayland Grades IV V VI VII VIII IV V VI VII VIII 130+ 1 1 125-129 120-124 1 115-119 1 110-114 1 105-109 1 1 100-104 1 1 3 1 95-99 1 90-94 1 1 1 1 85-89 2 2 2 1 80-84 75-79 3 2 4 1 70-74 1 2 65-69 3 1 1 1 1 60-64 2 3 3 1 2 1 1 55-59 2 1 1 50-54 2 3 3 1 1 1 45-49 1 2 2 1 2 40-44 3 2 2 1 1 2 2 4 35-39 3 3 1 1 1 3 30-34 3 2 3 2 1 4 1 25-29 1 1 20-24 3 2 1 15-19 1 1 1 2 10-14 1 1 5-9 2 1 1 0-4 No. Cases 17 32 16 13 17 11 12 11 12 15 Median 39 61 50 76 79 24 45 41 43 43 Standard 29 46 34 55 67 29 46 34 55 67 38 A Comparison of the Intelligence and Training of TABLE 9b. PEET-DEABBOBN PBOGBESS TESTS IN ARITHMETIC Average Operations May, 1922 Cochituate Wayland Grades IV V VI VII VIII IV V VI VII vi: 130+ 125-129 120-124 115-119 1 110-114 1 105-109 1 100-104 95-99 1 1 90-94 1 85-89 2 1 2 2 3 80-84 • 1 1 3 2 1 75-79 2 2 70-74 2 1 3 1 1 65-69 3 2 1 2 1 1 60-64 6 2 1 1 55-59 1 1 2 2 1 1 50-54 2 4 4 2 1 1 1 45-49 1 4 1 1 1 1 1 1 40-44 5 2 1 3 1 1 3 2 35-39 3 1 2 1 1 1 1 2 2 30-34 1 3 1 1 2 2 2 2 25-29 1 2 2 1 20-24 2 1 1 1 2 1 1 15-19 10-14 1 1 1 1 5-9 0-4 No. Cases 17 32 16 13 17 11 11 11 12 15 Medians 41 63 50 61 76 38 66 43 38 44 Standard 30 47 34 53 65 30 47 34 53 65 School Children in a Massachusetts Town 39 BESULTS OP THE READING TEST The Picture Supplement Test for silent reading is published in sev- eral forms of equivalent value, and as Form 1 was used in 1921, Form 2 was chosen for the 1922 testing. The disti-ibutions and medians will be found in Table 10. It should be remembered i'n connection with these figures that the scoring of this test is arranged to give a median score of 50 to the avea-age or normal class, irrespective of the grade. In the Third and Fourth grades the differences are not very great, and are slightly in favor of the Cochituate classes. In the Fifth and Sixth grades, however, Wayland shows decidedly to advantage. Cochituate medians are higher in the Seventh and Eighth grades. TABLE 10a. BURGESS PICTURE SUPPLEMENT SILENT READING TEST Distribution of Scores May, 1923 Cochituate Wayland Grades III IV V VI VII VIII III IV V VI VII VIII 100 1 2 98 92 1 86 2 1 80 1 1 1 1 , , , 1 74 1 1 1 1 1 68 2 1 2 3 2 4 3 3 1 1 62 1 2 3 3 3 2 1 2 56 1 3 3 1 2 2 1 1 1 3 50 5 1 5 1 2 44 3 4 4 1 1 2 2 38 6 2 5 2 2 2 5 1 1 2 32 1 3 4 2 1 1 1 1 1 3 20 2 7 3 1 2 1 1 1 2 2 20 2 3 1 1 2 2 1 14 1 1 8 1 1 ^1 1 1 No. Cases 24 16 34 16 13 15 12 11 12 11 12 15 Median 50 47 40 53 43 55 47 41 62 71 32 43 40 A Comparison of the Intelligence and Training of These results seem to imply some differences in the methods of teaching reading or the subject matter used or both. Further, the dif- ferences seem to exist not only between the two schools, but between the different grades in each separate school. RESULTS OF THE PENMANSHIP TEST The Holmes Test was used, exactly as in the previous survey, and the children were rated for both Speed and Quality of writing. The scores for Speed in letters per minute are shown in Table 11a. The Cochituate medians in every grade except the Eighth are higher than the Wayland medians. The differences, however, are hardly great enough to be of significance except in the Seventh grade. The distributions of the scores in Quality of penmanship are shown in Table 12a. The differences in the grade medians of the two schools are so small as to be insignificant, excepting possibly the Third grade, where the difference is five points, or half a step on the Ayres scale, in favor of Wayland. In neither school is there any improvement in the Quality of the writing after the Sixth grade. School Children in a Massachusetts Town 41 HOLMES TEST TABLE 11a. SPEED OP AVRITING May, 1922 Cochituate Wayland Grades II III IV V VI VII VIII III IV V VI VII VII 125-129 120-124 115-119 2 1 2 110-114 1 2 1 105-109 1 1 1 3 2 100-104 1 2 2 2 1 4 95-99 2 6 2 2 2 1 1 90-94 7 2 4 4 1 4 85-89 4 2 5 2 1 1 80-84 1 1 2 2 2 6 75-79 , . 1 4 2 2 1 70-74 1 1 2 3 1 2 3 65-69 1 3 3 4 1 1 1 60-64 1 2 1 55-59 1 1 4 2 1 1 50-54 1 4 6. 1 1 2 1 45-49 2 8 2 1 3 40-44 1 4 4 35-39 3 1 1 2 1 30-34 2 1 1 25-29 1 1 1 20-24 1 15-19 10-14 5-9 JNO. Cases 11 24 17 34 16 13 17 12 10 11 11 13 15 Med'n 46 48 54 80 99 94 96 42 46 74 91 ,82 101 42 A Comparison of the Intelligence and Training of HOLIVIES TEST TABLE 12a. QUALITY OF WRITING May, 1922 Cochituate Wayland Grades II III IV V VI VII VIII III IV V VI VII VIII 95-99 90-94 85-89 80-84 75-79 70-74 65-69 1 60-64 1 55-59 50-54 2 13 2 45-49 33122 102 40-44 34303 1 3222 35-39 374 8 4 2233 2 30-34 13374 3 3 2513 5 4 25-29 6 13 662 11 314112 20-24 4827 1231 1 15-19 10-14 5-9 No. Cases 11 24 17 34 16 13 16 12 11 12 11 13 15 Med'n 26 28 32 33 38 37 39 33 31 35 38 36 41 COMPARISONS WITH STANDARDS 1. The Arithmetic Tests In the following Tables 13a to 17a inclusive will be found the median arithmetic scores of both schools in comparison with the stand- ards. TABLE 13a. PROBLEMS May, 1922 V 43 45 38 The Fourth grade at Cochituate and the Eighth grade at Wayland are the only grades which fall below the standards in the Problem Test. In the other grades the medians are pretty uniformly well above the Grade IV Cochituate 13 Wayland 24 Standard 24 VI VII VIII 52 52 91 33 40 44 22 41 56 School Children in a Massachusetts Town 43 standards. It is worthy of note that there has been a decided improve- ment in the work at Wayland, for the results of the first testing sliowed not a single median which equalled or surpassed the standards in this test. TABLE 14a. ADDITION TEST May, 1922 Grade IV V VI VII VIII Cochituate 31 62 49 54 69 Wayland 24 64 39 38 48 Standards 30 50 35 52 60 In Addition the record is again good. The Cochituate medians are all higher than the standards, as are the Wayland Fifth and Sixth grade medians. In the Fourth grade the Wayland median is somewhat below the standard. The Seventh grade shows no progress over the Sixth, and the Eighth grade also is low. TABLE 15a. SUBTRACTION TEST May, 1922 Grade IV V VI VII VIII Cochituate 38 58 52 49 68 Wayland 51 65 63 25 34 Standards 28 47 39 57 66 In Subtraction the Cochituate medians are well above the standards except in the Seventh grade, where both schools fall below. The Way- land Eighth grade attained a median which is only about half what may be expected from the average class. The Wayland record is somewhat better than the previous one, as this time there are three instead of two classes above the standards. It is also noticeable that in the Fourth, Fifth and Sixth grades the Wayland medians are well above those of the Cochituate classes. TABLE 16a. MULTIPLICATION TEST May, 1922 Grade IV V VI VII VIII Cochituate 44 75 60 76 83 Wayland 42 83 39 30 51 Standards 28 48 31 51 68 In the Multiplication Test the Cochituate pupils obtained especially high scores, as the medians show. We again find the Wayland medians surpassing the standards in the first three grades, but they fall below in the Seventh and Eighth. The records are not greatly different from 44: A Comparison of the Intelligence and Training of those made at the previous testing except for the fact that the Cochit- uate Fourth grade and the Wayland Sixth grade have much better rec- ords in the 1922 tests. TABLE 17a. DIVISION TEST May, 1932 Grade IV V VI VII VIII Coehituate 39 61 50 76 79 Wayland 24 45 41 43 41 Standard 29 46 34 55 67 The Cochituuate class medians are all well above the standards in the Dix-ision Test. In Wayland the class medians are practically equal to the standards in the Fourth and Fifth grades, well above in the Sixth grade, biit very much below in the Seventh and Eighth grades. The chief change from the previous year is that the Wayland Sixth grade is now well above the standard instead of below it as formerly. TABLE 17b. AVERAGE OPERATIONS May, 1922 Grade IV V VI VII VIII Coehituate 41 63 50 61 76 Wayland 38 66 43 38 44 Standard 30 47 34 53 65 When the averages of the pupils' scores in the four fundamental operations are considered, the work of the Coehituate classes is well above the standard in every grade. This also applies to the Fourth, Fifth and Sixth grades in Wayland. In the Wayland Seventh and Eighth grades, however, the medians are considerably lower than the standards. SUMMARY OF THE RESULTS IN THE ARITHMETIC TESTS In all but two cases, the Coehituate medians in the Arithmetic Tests surpass the standards. Usually the difference in favor of Coehituate is a very large one, and many of the classes equal or exceed the standard average median of the next higher grade. This raises the question as to whether there is not too much emphasis on the teaching of Arithmetic in the Coehituate school. It is probable that the Coehituate children are, as the Intelligence Tests indicate, somewhat superior in native ability to children in many other localities, and so should be able to do better work. However, there is a point beyond which drill and prac- tice in arithmetic is not economical. It is quite certain that no addi- tional time or attention need be given to arithmetic in this school, and it might possibly be well to lighten the requirements. School Children in a Massachusetts Town 45 The results at Wayland are not quite so consistent. In genei'al, the medians of the Fourth, Fifth and Sixth grades practically equal or exceed the standards. In the Seventh and Eighth grades, however, the standards are considerably higher than the median performances of the "Wayland classes, with the exception of the Seventh grade median in the Problem Test. In general it may be said that the Wayland medians are below tho.se of the Cochituate classes, though there is an exception in the Subtraction Test. It does not seem likely that the low results in the Seventh and Eighth grades at Wayland are to be explained on the ground of inferior native abilitj' on the part of the pupils, as the Intelligence Tests show these pupils to be on practically the same mental level with the pupils in Cochituate. The difference may be due to differences in the efficiency of the teaching, or to dift"erences in the courses of study. As the tests and standards had been re^dsed since the pre\'ious test- ing, it is possible to make only very general comparisons with the work in 1921. It may be said, however, that Cochituate maintains its pre- vious liigh level of work, while conditions at Wayland have improved in that there are now three grades obtaining medians above the standards rather than only two as formerly. A very interesting point is brought out by reference to the study of the results of the Intelligence Tests. In 1921 it was pointed out that the Wayland Sixth grade could not be expected to attain the standards in any subject because the Intelligence Tests showed the mental level of the pupils in that grade to be rather low. In the second testing, how- ever, it was foimd that the new Sixth grade was a decidedly superior one, and as has been shown above, the Sixth grade medians were this time above the standards. This illustrates very well the desirability of securing .some concrete evidence on the mental make-up of the pupils who take subject matter tests. It shows the injustice of evaluating the work of a teacher on the actual accomplishments of the class after she has taught them without taking into consideration the kind of material she had to work with. 2. The Reading Test The scoring of this test is so arranged that a standard percentage distribution of scores is offered to which the performance of any srade from the Second to the Eighth inclusive may be compared. This stand- ard distinbution is shown in Table 18a, together with the percentage distribution of the scores made in each grade in the two schools. It is not to be expected that small classes like these will give smooth distribu- tions, but if a class has average ability half the cases or 50 per cent, should obtain scores above 50. Another way of comparing these scores 46 A Comparison of the Intelligence and Training of with the standards is to examine the medians of the several classes. The method of standardization is such that the average class should have a median score of 50. TABLE 18a. PICTURE SUPPLEMENT SILENT BEADING TEST Form 2, May, 1922 Per Cent, of Pupils Receiving Each Mark Cochituate Waylaud Grades III IV V VI VII VIII Standard III IV V VI VII VIII 100 8 18 98 1— 92 7 1+ 86 8 8 3 80 3 8 7 4 9 7 74 4. 6 7 6 9 8 68 8 6 6 19 13 26 8 25 27 8 7 62 4 13 9 19 10 25 17 9 13 56 4 19 9 6 15 11 17 9 8 8 20 50 20 3 33 12 8 17 44 13 25 12 6 7 11 17 18 38 25 13 15 13 15 10 17 46 8 9 13 32 4 19 12 13 8 7 8 8 9 8 20 26 8 20 23 7 6 17 9 8 9 17 13 20 6 19 8 4 9 17 17 7 14 3 3 8 8 6 3 1 + 2 1— 8 9 No. Cases 24 16 34 16 13 15 12 11 12 11 12 15 Median Score 50 47 40 53 43 55 50 47 41 62 71 32 43 According to these methods of comparison all the grades in Cochit- uate except the Fifth and Seventh are up to the standard in this test. In Wayland the Third grade practically reaches the .standard, and the Fifth and Sixth grades surpass the standards for these grades by con- siderable margins, but the other three grades fall much below. The score in the Seventh grade is an especially low one. School Children in a Massachusetts Town 47 There has been a marked improvement in the reading since the pre- vious testing, as shovm by the following comparison of median scores. Cochituate Wayland Grade III IV V VI VII VIII III IV V VI VII VIII 1921 38 56 20 25 26 32 26 32 38 20 20 38 1922 50 47 40 53 43 55 47 41 62 71 32 43 In 1921 the only grade median which equalled the standard was that of the Fourth grade at Cochituate, while the medians in the other eleven grades were extremely low. In 1922 only five of the medians were below the standard, and even in these five grades there has been a marked im- provement. This change alone would seem to justify the first survey which pointed out the deficiency in reading ability. 3. The Penmanship Test In Table 19a are shown the medians of the various grades in the Speed of Writing test, together with standards and the results obtained in three other Massachusetts communities. It will be seen from the table that the Cochituate medians are, except in the Fourth grade, some- Avhat higher than the Standards. They also compare very favorably witli the results obtained in the three systems listed in the table for com- parative piu-poses. The Wayland medians are slightly below the stand- ards in grades II, V and VII, and somewhat above in grades VI and VIII. The Fourth grade median in this school is the only one which is materially below the standard. In comparison with the other three school systems, Wayland shows well in grades V, VI and VIII, but falls considerably below them in grades III, IV and VII. TABLE 19a. SPEED OF PENMANSHIP May, 1922 Grade Ill IV V VI VII VIII Cochituate 48 54 80 99 94 96 Wayland 42 46 74 91 82 101 Standard 49 62 76 87 90 93 Newton 55 59 73 85 94 102 Brookline 76 87 90 Fall River 86 89 48 A Comparison of the Intelligence and Training of A comparison of the work iu the two years is shown below. Coehituate Wayland Grade III IV V VI VII VIII III IV V VI VII VIII 1921 60 61 77 85 100 113 48 58 68 63 88 98 1922 48 54 80 99 94 96 42 46 74 91 82 101 The losses are slightly more numerous than the gains, although there are no large losses except in the Third and Eighth grades at Coehituate. The Sixth grade in each school shows a rather large gain — 14 letters per minute at Coehituate, and 28 letters per minute at Wayland. The lat- ter gain is probably explained when we remember that the Wayland Sixth grade in 1921 was shown to be somewhat inferior mentally. Table 20a shows the results of the Quality Test compared with the standards and the results from the three school systems whose medians were used for comparison in the Speed Test. No median in any grade of either school equals or exceeds the standards of the results of the other systems. In practically every case the difference is a large and signifi- cant one. TABLE 30a. QUALITY OF PENTVIANSHIP May, 1928 Grade III IV V VI VII VIII Coehituate 28 32 33 38 37 39 Wayland 33 31 35 38 36 41 Standard 36 39 44 46 47 49 Newton 50 45 48 51 50 53 Brookline 44 46 47 49 Fall River 44 47 Neither school has done as well iu this test as in 1921, as may be seen from the comparison of the grade medians for the two years shown below. Coehituate Wayland Grade III IV V VI VII VIII III IV V VI VII VIII 1921 39 46 43 45 50 53 32 40 60 48 50 57 1922 28 32 33 38 37 39 33 31 35 38 36 41 Many of the losses are large, amounting to more than a full step on the Ayres scale, by which they were graded. School Children in a Massachusetts Town 49 SUMMARY AND SUGGESTIONS As the previous report pointed out, the results of such a survey as this must be used, in the main, by the teachers and administrative offi- cers who are on the ground and have full knowledge of the local condi- tion. However, there are certain points brought out in the preceding pages which may be commented upon by the surveyors before closing this report. In the first place, it will be noted that there are still many differ- ences between the performances of the children of the two schools in the various tests. When the first report was written it seemed possible to explain these differences largely on the basis of differences in native intelligence, but the 1922 testing seems to show that these variations were largely removed by the reorganization of the classes. It appears that some other reason must be sought for the differences now existing. Generally speaking, it may be said that the work of the Cochituate children is superior to that which is done at Wayland, although there are exceptions to this general rule. The Cochituate medians are also generally equal or superior to the standards in the various tests, except in the Quality of Penmanship. The arithmetic scores are so far above the standards as to indicate the possibility that some of the time spent on this subject could be used with greater profit in other fields. The Wayland results are an improvement over the previous year in arithmetic and reading. The Sixth and Seventh grades, however, are still below the standards in most cases. Both schools made marked gains in the Reading Test. In 1921 only one grade median equalled the standard, while in 1922 only five grades fail to attain the standards, and each of these had made some improve- ment over the previous performance. The results of the Penmanship Tests were not as satisfactoiy as those of 1921, since there were losses in both Speed and Quality. The Quality scores are decidedly too low, and efforts should be made to im- prove the handwriting. This should not be done at the sacrifice of speed, as the Speed scores are not too high. There are many exceptions to these general statements both among the grades and among the individual pupils. These cases must be singled out by superintendent and teacher for the special attention which they need. APPENDIX In order that the teachers might take full advantage of the results of the survey, tables like the following were appended showing the score of the individual pupils in each of the tests. A study of such a table by the teacher enables her to proceed intelligently in her efforts with individuals of her class. < SurDB9H oo'^iMOTtiiMcaooTtHcoo oocM^cn ''•fl-Y (N(MCOCCI(N(MCOCOM(MNCO COCOTtHIMCO •nriTrr toooiOcocococococDioo ocoococo "S D98dQ I-HOIOCOCCCOCDOOOOCO OOrHinOCO a) 3 S •3A'W OOOOOCD[~Lnt~CO'*COCO 00t:~i-HO3{M aj rfA W "^ ^'*~' ^^-^ '^ *"*^ >- '^•^ L"" ^'.1 "-T^ ^'^ '•'•' Q. V o5cx)ooajG5ooa5aooa30T-i '^ frt 1-1 T-H QJ m »! Q, 'UllU ^ OO-rHlOC^i-lt-OlOOCOIMC-'r'OOcnoOCO .5^^ • '^ a5C5COt-OOOa}OOCOt-OiOlSt~COt~<^t^ "'-' T-1 & "-I TO "- .'" L 00000000000000>-iiHCqiSC^O.-lrHtO 1—1 rHrH^Hi— (r^ i— ll— I t"' CO LO _ CO 00 IC 00 CO L-5 * ^ ^ '~' MSdo 'OAV c-iTftO(McoCT>cdiricoicocoo6 i^'ooscoio 5^ c^coooooTttint-mTht-t- ocoocDoo O "^ *'! -nTriTiT OCllOClOTtlOi-IOLOCOCO tH(MCO00 qng ocot~->*comiooom"*ioco t~cci-iiraoo I—* ."S -AT/T T-iTjHi-HincjCJCDt-asiMt-ai loco-ooci "yr fc4 AIQ cOTjiOinC-COtnt-COCOaDOD O^DCOLOCO <^ J •r>T-i-cr (r;'MC5IMCOrtcot-i--*cO'*t-'*ccooocoLOco ^ »— < I— ' t> •nn-in ■^■*woo«o^oocDO-*to-*T)HOoo-*co |> 4"'^<1 aOlMt-«D-*lOOO"*(MuOCv]m'*l:~Ttlt^(Mm ~Q-rT o o o o o o o ^•°Y -lOOOi-l uot^BuitnBxg 80n8St|I9?"I 92y 9raB|^ C~jt:~'— lt-COCOC3COOt— (M05C3O5O5C0C35m ■^i-ICOi-li-ICOOlCOcM^COTtHIMCvli-ILOi-IIM O 1— I O O iH LOl— I^H05i— iT-Hi-H(MCOOCOIMtDOOOOrHlO OOOT-HC5rHrH^(>'](r^lO'— II— lOrHi-ICItH M p^ M K ^' ^ W P^ ^ O ^' ^ W ^ '^ ^ "^^ ^ LIBRPIRY OF CONGRESS 019 842 692 1 1