LA 339 INS ( l A 436730 F p P I -o 4 rH*:I r+ iv. p -. + p,-< s^. 0 z t..;. *.?: F ,i I I I.:.% I-,' I-.?,A A I Y I rr- 1. ^^ ay"::/ii ''.:" S=J l-wS;M? irj- *21@4.,; i, k '.... -..i *~.:.::.. ~u-;..,,5,.-:.,. \..,. s. X..! ^';.:.,,/::, i-~,~..',::..a,*.< t A - ^. '.L^ i4 S^. V;;; - r.:,,, 1;.. '. '., r ':'... "'. 6 'Wi 0r^* v? t-*i-'.( ^ *..'^' -*R-^. I- "'w ^ / ^ iL} oEl-i"?( ) - -tS ) " "A Irllllll~ ll111 [111111 RECEIVED IN EXC HANGE RFROMJ I I! viQif L. < z:: n lau~w~~nal iiu ~lulun I m m ~~~~lHlu~jlhlilululuL!, IV T . I, VI' 0 0 I., 1. —? (x I A, N- 49 ~ 4 I~~~~~~~~~~~~~~~~~~~~~~~~~~~~i ie A,~~~~~4;~ v w,~~~~~~~~~~~~~~~~~ ( 0-3~~~~~~~~~~~~~~~~~~~4 p Av44 I PART-TIME TYPES 3LsOF ELEMENTARY SCHOOLS IN NEW YORK CITY A COMPARATIVE STUDY OF PUPIL ACHIEVEMENT BY. E FRANK M. J3UANCE SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN THE FACULTY OF PHILOSOPHY, COLUMBIA UNIVERSITY., BUREAU OF PUBLICATIONS Teacber. College, Columbia lLniber0itp NEW YORK CITY 1926 Copyright, 1927, by TEACHERS COLLEGE, COLUMBIA UNIVERBITY Printed in the United States of America by J. J. LITTLE AND IVES COMPANY, NEW YORK CONTENTS CHAPTER I. THE PROBLEM.......... II. THE DATA, METHOD AND RELATED STUDIES. III. THE SCHOOLS-THEIR ORGANIZATION.. Selecting the Schools....... School and Class Enrollment The School Program.... The Time Schedule. Assistants to the Principals. Methods of Promotion Defective Classes......... IV. THE SCHOOLS-THE TEACHERS AND PUPILS 1. The Teachers......... Professional Training.... Experience... Attitude of Teachers..... 2. The Pupils..... Birthplace of Parents..... Age of Pupils.... Attendance........ V. PUPm ACHIEVEMENT.... The Tests........... PAGo......... 11..... 4...... 11..... 121....,....13,........15.........17....,....18......... 18 ~ ~.....~ ~ 20 ~ O..... ~ ~ 20 ~ ~..... ~ ~ 20....2....23 ~........ ~ 24.... 25....,....25 *~ X ~ * 26 ~...........27...........29........... 29 Statistical Treatment to Secure Comparable Measures of Differences............ The F Technique...... The Equivalent-Groups Method......... Lower and Upper Halves of the Equated Groups.. VI. SUMMARY AND CONCLUSIONS............ BIBLIOGRAPHY.............. 30 32 36 42 45 49 I TABLES NUMBER PAGE I. DISTRIBUTION OF PUPILS AMONG THE GRADES OF THE EXPERIMENTAL AND CONTROL SCHOOLS......... 5 II. SCHOOL AND CLASS ENROLLMENT IN THE EXPERIMENTAL AND CONTROL SCHOOLS........... 12 III. TIME SCHEDULE IN MINUTES PER WEEK FOR THE EXPERIMENTAL AND CONTROL SCHOOLS IN ARITHMETIC, ENGLISH, SPELLING AND DICTATION, AND PENMANSHIP..... 16 IV. MINIMUM AND MAXIMUM TIME SCHEDULE OF THE NEW YORK CITY SCHOOLS IN MINUTES PER WEEK.. 16 V. YEARS OF ACADEMIC AND PROFESSIONAL EDUCATION OF TEACHERS 21 VI. THE PRE-SERVICE PROFESSIONAL TRAINING OF THE TEACHERS IN THE SCHOOLS.............. 21 VII. COURSES TAKEN IN THE IN-SERVICE PROFESSIONAL TRAINING OF THE TEACHERS WITH NUMBER IN EACH TYPE OF SCHOOL TAKING THEM............ 22 VIII. TOTAL EXPERIENCE OF THE TEACHERS........ 24 IX. BIRTHPLACE OF THE PARENTS OF THE PUPILS OF THE SCHOOLS 26 X. PERCENTAGE OF PUPILS OF NORMAL AGE, UNDER-AGE, AND OVER-AGE IN THE THREE TYPES OF SCHOOLS..... 27 XI. ATTENDANCE OF PUPILS DURING 233 SCHOOL DAYS IN THE THREE TYPES OF SCHOOLS......... 28 XII. F SCORES BY GRADES IN THE THORNDIKE-MCCALL READING SCALE FOR THE THREE TYPES OF SCHOOLS..... 33 XIII. F SCORES BY GRADES IN THE BURGESS SCALE FOR MEASURING SPEED IN SILENT READING FOR THE THREE TYPES OF SCHOOLS 34 XIV. F SCORES BY GRADES IN THE WOODY-MCCALL MIXED FUNDAMENTALS IN ARITHMETIC FOR THE THREE TYPES OF SCHOOLS 35 XV. F SCORES BY GRADES IN THE SPELLING TEST FOR THE THREE TYPES OF SCHOOLS........... 36 XVI. F SCORES BY GRADES IN THE HANDWRITING (QUALITY) TEST FOR THE THREE TYPES OF SCHOOLS....... 37 XVII. F SCORES BY GRADES IN THE HANDWRITING (SPEED) TEST FOR THE THREE TYPES OF SCHOOLS........ 38 vii viii Tables NUMBER PAGE XVIII. SUMMARY OF TABLES XII-XVII........ 39 XIX. MEAN T SCORES IN THE TESTS FOR THE EQUATED GROUPS FROM THE THREE TYPES OF SCHOOLS, THE DIFFERENCES BETWEEN THE MEANS AND THE PROBABLE ERROR OF THE DIFFERENCES.............. 41 XX. MEAN T SCORES IN THE TESTS FOR THE UPPER AND LOWER HALVES OF THE EQUATED GROUPS FROM THE THREE TYPES OF SCHOOLS, THE DIFFERENCES BETWEEN THE MEANS AND THE PROBABLE ERROR OF THE DIFFERENCES..... 43 ACKNOWLEDGMENTS In a study which was made possible only by the cooperation of so many, it is difficult to make acknowledgment to all to whom I am indebted, but my thanks are specially due to the principals of the schools in which the investigation was conducted, viz., P. S. 53 Bronx, P. S. 7 Bronx, P. S. 103 Brooklyn, and P. S. 106 Brooklyn. To them and to their assistants and teachers I owe much for their splendid cooperation. I desire to record my thanks also to Mr. Hugo Newman, principal of the New York Training School for Teachers, for permission to secure the assistance of senior students from the Training School in connection with the giving of the tests in the schools. Finally, I am deeply grateful to the members of my committee -Professors A. I. Gates, R. Pintner, W. A. McCall, and E. S. Evenden for their friendly assistance and helpful criticism throughout the study. F. M. Q. PART-TIME TYPES OF ELEMENTARY SCHOOLS IN NEW YORK CITY CHAPTER I THE PROBLEM LOCAL SITUATION The present varied character of the organization of the intermediate grades of New York City's elementary schools is due more to the extraordinarily rapid growth of the city than to any other one factor. It is true that for a few years following the latter part of 1914 an experiment was tried with the Gary system of school organization. This system which is based upon a theory of curriculum construction only incidentally helped to meet overcrowded conditions, and was discontinued because of opposition developing in certain sections. A decade earlier than this, however, the congested condition of the schools was being acutely felt. As evidence, we find a letter in the Journal of the Board of Education of September 21, 1904, written by Mayor McClellan to the Board of Education in which he submitted a plan by which the school facilities, with the exception of the auditoriums, were to be more fully used to obviate the necessity of placing pupils on part-time. On September 10, 1913, a Special Committee on Part-Time Conditions, which had been appointed by the Board of Education, brought in its report. After indicating a number of devices, considered unsatisfactory by them, by which part time might be reduced or avoided, the Committee outlined six systems, or programs, for part-time classes which were in use. Because of the direct bearing of the first and fourth of these on the present study they will be given here as outlined by the Committee. 1 2 Part-Time Types of Elementary Schools FIRST SYSTEM Morning Class............................. 8.30 to 12.15 Afternoon Class........................... 12.30 to 4.15 Comparison of First System with Full Time: Full Time: Length of school day.......................... 6 hours Time in school................................ 5 hours Lunch recess................................... 1 hour Part Time: Length of school day........................... 3 4 hours Time in school.............................. 3% hours Objections: Children on the street either morning or afternoon; Strain of a long, single session; Children are taught in the long session at the time of waning activity. FOURTH SYSTEM Early-Time Classes: 8.30 to 10.30..........Classroom 10.30 to 11.00.........Playground 11.00 to 12.00..........Luncheon recess 12.00 to 12.45..........Use of room of full-time class, then at luncheon recess. 12.45 to 2.00......... Classroom 2.00 to 2.30..........Early-time teacher assists late-time teacher Time in classroom.....4 hours Time in playground.... Y hour Late-Time Classes: 10.00 to 10.30..........Playground 10.30 to 12.45.........Classroom 12.45 to 1.45.......... Luncheon recess 1.45 to 2.00..........Playground 2.00 to 3.45..........Classroom 9.30 to 10.00..........Late-time teacher assists early-time teacher Time in classroom 4 hours Time in playground y4 hour Comparison of Fourth System (Strachan Modification) with full time: Full Time: Length of school day............................. 6 hours Time in school................................... 5 hours Lunch recess................................... 1 hour Strachan Modification Group I Group II Length of school day....................... 5/2 hours 5% hours Time in school............................ 41/2 hours 4% hours Lunch recess............................... 1 hour 1 hour The Problem 3 When an auditorium is available, the Fourth System outlined above undergoes further slight adjustments to enable the teachers and pupils to benefit thereby. But these two systems with their modifications along with the traditional or regular session constitute the systems under which the majority of the schools of the city at present operate. THE PROBLEM The problem which forms the basis of this study was suggested by the Problems Committee for 1921-1922 of the Principals' Association of the New York City Elementary Schools, under the chairmanship of Dr. J. K. Van Denberg. To them the question of part-time schools was a crucial and pressing one in the city. They knew that there was a variety of methods of school organization by which an attempt was being made to solve tentatively the problem of overcrowding. But as to what extent, if any, the efficiency of instruction in the fundamentals of the curriculum was being affected by these adjustments in organization they did not know. Hence, they decided to select this problem for study under their auspices during the year, and the writer was approached with a request to conduct the investigation for them. The problem resolved itself finally into the answer to two questions: 1. Selecting two representative types of part-time schools, in what essential features do they differ in organization from those of the traditional or regular session school? 2. Taking into consideration all the important relevant factors, and using the regular session school as the control school, in which do the pupils show the best results of instruction in the fundamental subjects of the curriculum? It should be noted at the outset that the problem was circumscribed by important limits, that not all the subjects of the curriculum came within the scope of the investigation, and that the effects of the respective systems of organization upon such important matters as general information, health, attitudes, and desirable qualities of citizenship were of obvious necessity unmeasured, though their vital importance was clearly recognized. CHAPTER II THE DATA, METHOD, AND RELATED STUDIES THE DATA The two types of part-time schools selected were the morning and afternoon, or two-session type, and the duplicate, or that resembling the fourth system reported by the Special Committee on Part-Time Conditions outlined above. These hereafter will be referred to as Experimental Schools and will be distinguished from each other by being designated, respectively, the TS and the DS types of schools. The regular session, or Control School, will be referred to on occasion as the RS school. The grades selected were 3B to 6B, inclusive. The work in these grades is not only representative of that of the whole of the elementary school, but being below the junior high school grades is based upon a uniform course of study throughout the city. Except in the case of the control school, the study was confined to one school of each type of organization within the range of the grades specified above. Because of a comparatively smaller number of pupils in three of the grades in this school who were able to meet the conditions of selection, this number was supplemented by additional pupils from another control school. This second control school will hereafter be referred to as control school (1), or RS (1). Altogether, 1,917 pupils from all the schools were included in the initial study. For various reasons, such as absence on one of the days of the tests or because of data being incomplete in other respects, only 1,786 were retained in the final study. These were distributed among the different grades of the schools as shown in Table I. Data were gathered about the pupils, the teachers, and the organization of the school. In the case of the last two, this was done by personal interview and questionnaire. The data about 4 The Data, Method, and Related Studies 5 TABLE I DISTRIBUTION OF PUPILS AMONG THE GRADES OF THE EXPERIMENTAL AND CONTROL SCHOOLS Grades School - Total 3B 4A 4B 5A 5B 6A 6B TS............... 91 93 98 98 99 90 90 659 DS............... 95 91 93 76 92 88 86 621 RS............... 43 75 74 63 57 45 77 434 RS (1)............ 25 24 23 72 Total............ 254 259 265 237 272 246 253 1,786 the pupils were collected by the teacher, or where given by the pupil, were checked by the teacher. Standard tests in the fundamentals of the curriculum were given on March 29 and 30, 1922, by forty senior students of the New York Training School for Teachers who had had a thorough course in tests and measurements under Dr. F. M. Hamilton, head of the department of psychology in the Training School. In addition, the writer spent two days with these students giving further training to insure that all should interpret alike the standard instructions accompanying the tests and administer the tests in strict accordance with those instructions. THE METHOD Circumstances permitted but one test to be given. Hence, to eliminate any transference of effects of other types of schools upon the subjects of the investigation, only those pupils who had been attending school for two years in the respective systems were included in the investigation. The main problem was to determine the relative amounts of achievement made by the pupils in the respective types of school organization as measured by the educational tests. But in addition to being able to compare the differences in gross scores between the groups in the same test, it is necessary to have some common unit in which these differences may be compared among different tests. A gross difference of 8, for example, in one test is not necessarily equivalent to the same difference in another test. The size of the initial and final scores in each may be different so that a common difference of, say 5, does not measure equal increments. 6 Part-Time Types of Elementary Schools To secure this comparable measure all the scores were transmuted into T scores which is a measure of the dispersion of the twelve-year-old group.1 That is, the differences in scores when thus measured are sigma differences and thus may be compared. The method by which the scores in the educational tests were thus transmuted will be explained in a later chapter. When this common T measure was found the data were subjected to two methods of statistical treatment. 1. The F 2 Technique. By this technique the averages of the scores in the educational tests of all the pupils in each group were compared and interpreted in the light of the average scores in the intelligence test. It is obviously unfair to expect that group with the lowest mental ability to secure the highest scores in the educational tests. By this method any such disadvantages are equated. 2. The Equivalent-Groups Method.8 Here an attempt was made to make a finer study of the data. Pupils were equated, wholly or within very narrow range, with respect to seven variables. These variables were sex, chronological age, intelligence, birthplace of parents, regularity of attendance, number of terms at school and grade. The differences in the educational scores of these equated groups were found and the reliability of these differences measured. This method will also be explained in more detail when this part of the study is being considered. RELATED STUDIES The development of the scientific instrument of objective measurement in education has taken the measurement of the results of teaching out of the realm of general impression. It has made it possible for educators in widely separated localities to measure amounts of educational achievement to be found in children of these localities, apd for purposes or comparison these measurements run no risk of being inaccurate through subjective determination. The availability of this instrument has been of inestimable service to the educational administrator. Innumerable surveys 'McCall, W. A. How To Measure in Education. Macmillan Company, New York, 1922, Chap. X. 2McCall, W. A. How To Expertment in Education. Macmillan Company, New York, 1923, p. 103. 'bid., Chap. VII. The Data, Method, and Related Studies 7 are witness to this fact. These have been carried on very extensively in urban communities with types of school organization closely resembling one another. The literature on this phase of educational measurement is very extensive. But programs of educational testing have been carried out for purposes of comparing different systems of school organization. Sometimes these have been confined to cities; at other times they have been extended to rural school systems. One of the earliest and most important of the former type was a study of the Gary schools at Gary, Indiana. Courtis' says of these schools, "As contrasted with education of this meager type, the Gary plan is distinguished by two features, intimately connected with each other: first, the enrichment and diversification of the curriculum; second, the administrative device that, for want of a better name, will be tentatively termed the duplicate school organization." Tests were given in these schools in handwriting, spelling, arithmetic, composition and reading, and in these tests the results were found to be poorer than in other cities. No account was taken of the mental ability of the pupils in these schools. In 1915, Buckingham, chief statistician at that time of the Department of Education of the City of New York, conducted an investigation the purpose of which he states as having been "to determine to what extent the Gary schools and the PreVocational schools have been able to give as effective instruction in the academic subjects as have the regular schools." Two tests were given in March and June to the seventh and eighth grades and the amount of improvement measured. Buckingham found that for all the subjects in which the tests were given the Gary schools yielded the least amount of improvement; and the regular session, the greatest. It should be noted, however, that this investigation was carried on only a few months after the organization of the Gary schools. This circumstance may have caused the poorer standing of this type of school in the study. After outlining the nature of the organization of the academic, 'Courtis, S. A. The Gary PubUlo Schools-Measurement of Classroom Products. General Education Board, 1919, p. ix. Seventeenth Annual Report of the iCty guperintendent of Sohools, 1914 -1915. Survey of the Gary and Pre-vocational Schools in New York City. 8 Part-Time Types of Elementary Schools industrial, and duplicate types of school in Seattle, Ayer 6 states "with the purpose of arriving at the most rational solution of the problem of organization... I have attempted to analyze the results which have accumulated under the several forms of organization operating in the city of Seattle." He says again, "A number of detailed studies have been made of the comparative efficiency of sixteen local elementary schools selected to represent the academic, duplicate, and industrial types of organization and compared upon a basis of accomplishment in reading, arithmetic, school progress, final elementary marks and high school teachers' marks." In conclusion he states "that the departmentalized duplicate and junior high schools have the better record of achievement when measured by tests in the elementary school subjects." The study was based on 324 graduates of the Grade VIII classes in the three types of schools. In a thoughtful and extensive study of the platoon schools in Detroit, Spain7 represents parents and educators as asking, "Is the platoon school producing satisfactory results in the three R's?" To answer this question he takes the results of a testing program extending over four years and compares the results from platoon schools with those from the city as a whole. In summing up, he says: "From these studies it seems reasonable to conclude that the platoon school after it has passed through the period of readjustment, may be expected to give more efficient training in the fundamental skills than the non-platoon schools have given." All the studies mentioned thus far have been of certain modifications of the traditional urban elementary school. But in another study, Frost8 collects the results of twenty-eight investigations in which comparisons between the rural and town schools had been made. In summary, he says, "In all, 28 studies are reported of which 20 find the results better in urban schools, 6 find negligible differences, and 2 find results superior in the rural schools." In his own study in which he carried on an extensive testing program, he found the town school superior Ayer, F. "The Curriculum-Organization." A mimeographed Report to the Board of Education, Seattle, 1923. 7 Spain, C. L. The Platoon School. Macmillan Co., 1924, pp. 166, 182. ' Frost, N. A. Comparative Study of Achievement in Country and Town Schools. Teachers College Contributions to Education, No. 111, 1921, p. 14. The Data, Method, and Related Studies 9 in its test results to both the 6-month and the 9-month rural school. This was particularly so in relation to the former. Foote9 as chairman of a committee of the Rural School Department of the National Education Association sought to find "in which of the two outstanding types of rural schools, the consolidated and the one-teacher, are the results of instruction superior?" After a very extensive study including more than 15,000 pupils in 509 schools in 19 states in which educational tests were given he finds that "the rate of progress from grade to grade is very nearly equal in the two schools; pupils advance about as rapidly in one type as in the other"; and "there is a significant difference in the results of instruction in each grade tested in favor of the consolidated school," but asks the question later, "Why is there such a small difference in the results of instruction between the two types of school?" No measure of intelligence was used in this study because of the assumption "that general intelligence in the two groups is similarly distributed." In the report of the Rural School Survey of New York State, Haggerty 10 summarizes the results of a testing program covering the one-teacher and the four-teacher rural schools of the state. The tests given were in reading, arithmetic, and history. The difference in the scores of the tests between the two types of schools was significant and indicated a consistent superiority of the pupils in the larger type of school over those in the oneteacher school. In conclusion, Haggerty says, "Whatever the detailed cause may be, however, the fact remains that the oneteacher school is a less productive educational institution than is the larger school unit, and the pupils who attend the smaller schools are being handicapped for life by this fact." A recent book by Collings 11 gives the results of a study in which the traditional rural school is contrasted with one whose organization is determined by the project method. He states the purpose of the study as being "to throw some scientific light upon the following questions: Can the country school curriculum be selected directly from the purposes of boys and * Foote, John M. "A Comparative Study of Instruction in Consolidated and One-Teacher Schools." The Journal of Rural Education, 2:8, 1923. 10 Haggerty, M. E. Rural School Survey of New York State. A Report to the Rural School Patrons, Ithaca, New York, 1922, p. 169. ' Collings, E. An Experiment with a Project Curriculum. Macmillan Company, New York, 1923, p. 4. 10 Part-Time Types of Elementary Schools girls in real life? If so, to what extent, with what effect, and under what conditions?" For the purpose of showing what the outcome of the curriculum was "in terms of common facts and skills," standard tests were given both in the experimental and control schools at the beginning and the close of a fouryear period. The equivalent-groups method was used. The findings favored the experimental type of school organization and curriculum throughout. In conclusion, it may be noted from the preceding survey of related studies that four of the eight mentioned are concerned with the comparison of the traditional and modified forms of urban school organization, with respect to their effects upon pupil achievement. In two of these four, the traditional type of school organization is reported as giving superior educational results. The four remaining studies report the findings of comparisons made in educational achievement of the pupils between the rural one-teacher school and each of the following in turn: the town school, the consolidated school, the larger rural school, and the rural school organized about a new type of curriculum. That type of rural school organization known as the traditional one-teacher school was found in each case to give the poorest educational results. CHAPTER III THE SCHOOLS-THEIR ORGANIZATION SELECTING THE SCHOOLS In February, 1922, Mr. Eugene B. Gartlan, the successor of Dr. Van Denberg to the chairmanship of the Problems Committee of the Elementary School Principals' Association, secured from the City Department of Education a list of approximately twenty schools of each type of school organization selected as the basis of this study. Thereupon, the following facts were asked by letter from the principals of each of these schools: 1. The number of pupils in each grade from 3A to 6B, inclusive, who are enrolled under the system in your school. 2. How many successive terms immediately preceding this one has the majority of the pupils in each of these grades been working under this system? 3. Do the majority of the pupils in these grades encounter language difficulties because of non-English homes? When replies had been received, the writer visited the schools with a view to securing such representatives from each type as were as nearly alike as possible, keeping in mind the necessary variations due to the different systems of organization. One of the important factors to consider was that of the language difficulties of the pupils. Another was the social and economic character of the communities. As to the administration of the school itself, an important factor was that of comparative size. Obviously, it was impossible to equate these and other factors with any exactness, but it was desired to keep the differences at a minimum. With these factors in mind P. S. 103 Brooklyn and P. S. 53 Bronx were selected as the representatives of the two types of the experimental schools, the former of the morning and afternoon session-or two-session type-and the latter of the duplicate type. The two schools, P. S. 106 11 12 Part-Time Types of Elementary Schools Brooklyn and P. S. 7 Bronx were chosen as the representatives of the regular session, or control type. The communities are typical of a population engaged in merchandising and various occupations of a manual character. From the economic aspect the parents of the children in Experimental School TS are in the most advantageous position, although in none of the school communities is there any evidence of poverty. In this community the occupations of the parents ranking highest in frequency are the manufacture of wearing apparel and the carrying on of some form of mercantile business. The parents of the pupils in Experimental School DS are also engaged in the manufacture of wearing apparel. Other parents are merchants, painters, plumbers, and salesmen. In the communities of the control schools the parents are for the most part merchants, mechanics, factory workers, plumbers, and builders. The range of occupations in all the communities is very wide, however, with a few professional men in each. SCHOOL AND CLASS ENROLLMENT In size, as measured by the number of classrooms, the schools that were selected for this study lie near the median of the schools throughout the city. Obviously, the number of pupils in the experimental schools would be greater than in the control schools where each classroom seat accommodates but one pupil. The facts with respect to enrollment are given in Table II. TABLE II SCHOOL AND CLASS ENROLLMENT IN THE EXPERIMENTAL AND CONTROL SCHOOLS Total Number Average Number of Pupils per Teacher School Enroll- of ______ ment Regular Teachers 3A 3B 4A 4B 5A 5B G6A GB TS........... 2,407 49 39 55 48 53 51 51 51 52 DS........... 3,013 71 49 45 43 43 41 44 38 41 RS........... 1,855 47 48 38 41 39 35 37 37 33 RS(1).........1,709 52 41 39 41 32 35 35 31 49 The total enrollment of the four schools is seen to be approximately 9,000 pupils. The figures giving the average number of pupils per teacher show a large enrollment in Experimental School TS, somewhat The Schools-Their Organization 13 less in Experimental School DS and still less in the Control Schools RS and RS (1). From many aspects it is highly desirable to keep a classroom enrollment down to a maximum of 35 pupils, but Stevenson found in his study of the relation of class-size to the efficiency of teaching that "the size of classes, even when important influences affecting teaching are made constant, seems to determine the result only to a very small degree." As the enrollments by grades in the four schools in this investigation fall almost entirely between 33 and 50, the range of class-size used by Stevenson in his study, the evidence points that the experimental schools are handicapped very little, if any, by the size of their classes. SCHOOL PROGRAM The school program for Experimental School TS is the same as outlined in Chapter I under the First System described by the Committee on Part-Time mentioned there. This was as follows: Morning Class....................... 8:30 to 12:15 Afternoon Class.....................12:30 to 4:15 In Experimental School DS the organization, though very similar to that under the Fourth System described by the same Committee, yet has certain modifications. The exact schedule is given here: Group A Group B 8:30-10:15 Classroom 9:30-10:15 Playground 10:15-11:00 Playground 10:15-12:00 Classroom 11:00-12:00 Lunch 12:00- 1:00 Lunch 12:00- 1:45 Classroom 1:00- 1:45 Auditorium 1:45- 2:30 Auditorium 1:45- 3:30 Classroom By this arrangement the length of the school day is the same as under the Fourth System but the length of time spent in the classroom is less by thirty minutes. There is an addition to the program, however, of an auditorium period of forty-five minutes. The total time spent at school by the pupils under the three systems is as follows: 1 Stevenson, P. R. Smaller Classes or Larger-A Study of the Relation of Class-Size to the Efficiency of Teaching. Public School Publishing Company, Bloomington, Ill., 1923, p. 69. 14 Part-Time Types of Elementary Schools T S.....................................33/ hours D S.....................................5 hours R S.....................................5 hours It will be noticed from the outlines of the above programs that in both of the experimental schools there are early and late classes. In School DS part of the pupils come at 8:30 in the morning while the others come an hour later. In School TS some of the pupils come in the morning and others in the afternoon. At regular intervals these early and late classes alternate. There are two outstanding features in the program of Experimental School DS. It is a direct result of the presence of these that the school is able to accommodate practically twice as many pupils as it could were it operating under the regular session system. One of these features is its daily playground period of 45 minutes. There are thus 225 minutes devoted weekly to the physical education of the pupils in this school. The weekly time allotment in all the schools to physical education is as follows: TS...................................150 minutes D S...................................225 minutes RS............................. 200 minutes RS (1)...............................200 minutes From these figures it will be seen that the difference is really significant only between the two types of experimental schools. The second outstanding feature of the program of Experimental School DS is its daily auditorium period of forty-five minutes. As in the playground period, so in this, the regular classroom teachers are responsible for the supervision of their pupils. In the auditorium there are gathered at the same time classes which are not always the most homogeneous with respect to grade. Sometimes they are as widely separated as two years in their academic standing. This precludes the possibility, from the viewpoint of the class teacher, of carrying on in the auditorium a type of academic work as profitable as it might be for the purpose of supplementing the regular class work of the pupils. And yet, that the time spent in the auditorium is not by any means lost in this respect may be seen from the nature of the activities which are there carried on. The Schools-Their Organization 15 1. Among the individual classes the following are some of the types of work given: Flash cards for studying arithmetical tables Study of weights and measures tables Phonics Memory gems Spelling Silent reading Geography when it is given as a lesson to be read from a book 2. From the platform are conducted the following forms of school activities with all the classes attending: Music Talks on manners and morals Talks on safety Correction of errors of speech Story-telling and reproduction Children describe flowers, fruits, vegetables, animals, objects in the room, etc., and the other pupils guess what is being described Contests in rapid calculation, tables, and spelling Language games In Experimental School TS there are no assembly periods. Both of the control schools have these periods but no classroom work as such is given in them. They are used for pupils' literary programs and similar activities. TIME SCHEDULE In 1913 there was introduced into the New York elementary school system a minimum and maximum time schedule. Its purpose was to give a little elasticity to the individual school in its distribution of time among the different subjects of the curriculum. In some schools, or in some classes within a school, special attention to a particular subject was needed. It often occurred, for instance, that the majority of a school or class enrollment came from a district the most of whose population came from non-English-speaking countries. In such a situation progress in the school work was largely dependent upon the 16 Part-Time Types of Elementary Schools acquiring of ability to speak and write the English language. This system permitted such extra time to be devoted to it. In Table III will be shown the average number of minutes devoted to arithmetic, spelling and dictation, English and penmanship in each of the experimental and control schools. In Table IV will be given the minimum and maximum time schedule in the same subjects as adopted by the Board of Education in the fall of 1921. TABLE III TIME SCHEDULE IN MINUTES PER WEEK FOR THE EXPERIMENTAL AND CONTROL SCHOOLS IN ARITHMETIC, ENGLISH, SPELLING AND DICTATION, AND PENMANSHIP Third Year Fourth Year Subject TS DS RS RSS(1) TS DS RS RS(1) Arithmetic....... 200 225 200 200 200 200 200 180 English......... 525 625 675 640 440 475 525 555 Spelling **........ Dictation........200 125 130 100 150 100 125 120 Penmanship...... 60 75 75 75 60 75 75 75 Fifth Year Sixth Year Subject TS DS e S RS(1) TS DS RS t S(1) Arithmetic....... 200 225 200 180 200 200 200 180 English....... 440 375 525 440 360 440 525 470 Spelling **........ Dictation.... 100 110 125 105 100 115 125 90 Penmanship...... 60 75 75 75 60 75 75 75 * The time allotment for reading is not given separately from that for English because the line was not sharply enough drawn between Reading and related material in the time schedules returned by the teachers. ** It should be noted that English includes Spelling and Dictation. The time for the latter has been deducted from the total indicated for the former. TABLE IV MINIMUM AND MAXIMUM TIME SCHEDULE OF THE NEW YORK CITY SCHOOLS IN MINUTES PER WEEK Third Year Fourth Year Fifth Year Sixth Year Subject Min. Max. Min. Max. Min. Max. Min. Max. Arithmetic........ 150 240 150 300 150 270 150 240 English...... 615 675 525 555 405 555 405 585 Penmanship 75 75 75 75 75 75 75 75 75I 75] 75I.5. 7 The Schools-Their Organization 17 From Tables III and IV the following points may be noticed: 1. The greatest uniformity among the schools in allotment of time is with respect to penmanship. 2. In addition to the time given to the mechanical phase of arithmetic in the auditorium periods, Experimental School DS assigns slightly more in the regular classroom to the subject than do the other schools. 3. Control School RS allots more time to English than do the other schools. A part of this extra time in the fifth and sixth years is given to spelling. 4. Because the minimum time schedule in all the subjects on the curriculum calls for more time than is spent in the classroom in the experimental schools, the minimum time is not always possible in every subject in these schools. In this connection it should be said that an attempt was made to ascertain the amounts of time that were given to homework by the pupils in the respective schools. The data as given by the parents of the children were found upon examination to be so unreliable that they were discarded. Homework was given in each of the schools, however, but there was no evidence that the pupils in any one of the schools spent more time than did those of another in this way. ASSISTANTS TO THE PRINCIPALS The number of assistants to the principal is determined by the size of the school. To schools of the size selected for this study two are allowed by the Board of Education. Their duties are both administrative and supervisory. In large schools such as these it is obviously impossible for the principal to interview all of the parents who come to the school from time to time. His assistant takes his place in connection with such interviews. Any other matter of a routine nature which affects the classes of which he has charge are referred to him. In a real sense he stands in the place of the principal. The principal, however, has general supervision over the whole school, and in particular, where time permits, over the senior grade of the school. He is responsible ultimately for the school's standards and progress. The chief duty, however, of the assistant is the supervision 18 Part-Time Types of Elementary Schools of the work of the classroom teacher. This is done by visits to the classroom, the teaching of demonstration lessons, and such subsequent work as is necessary to insure the acquisition of the desired method of teaching by the classroom teacher or the improvement of the class in any part of the curriculum. In none of the schools are the results of teaching measured by a carefully planned program of testing by means of the best standardized intelligence and educational tests. The method of determining the progress which the pupil makes from time to time is by means of examinations prepared in the individual school on the prescribed course of study. METHODS OF PROMOTION The system of promotion is virtually the same in all the schools. A record of the mark of each pupil for the month is kept on his own card. All deficiencies are reported to the parent each month on the pupil's term report card. The score throughout the term is cumulative so that deficiencies in the earlier months may be overcome by better records later in the term. The mid-term examination is set by the office; the final examination is set, however, by the individual classroom teacher, except in RS (1) where the monthly tests are set by the teacher and the final is set by the office. The office scrutinizes the number of promotions made by the room teacher to see that a satisfactory standard is maintained. This standard in the several schools in terms of percentage of classroom enrollment promoted is as follows: T S....................................90 per cent D S...................................85 per cent R S...................................85 per cent RS (1)...............................85 per cent DEFECTIVE CLASSES Classes for the feeble-minded are organized in only certain schools throughout the city to which from the regular classes in neighboring schools feeble-minded pupils are drawn off and sent. This does not imply that normally there is a greater percentage of such pupils in the school in which these classes are The Schools-Their Organization 19 held than elsewhere. Such is not the case. In only one of the four schools did it happen that there were any classes of this type. These were in Control School RS. In this school there were organized two such classes, one for boys, and another for girls, the total enrollment of the two together being 35. There were about an equal number of such pupils drawn off from all the schools, but as this number did not exceed eight in any one school, it was a negligible factor in so far as it affected the standing of the classes in this investigation. CHAPTER IV THE SCHOOLS-THE TEACHERS AND PUPILS 1. THE TEACHERS PROFESSIONAL TRAINING The preeminence of the teacher as a factor in bringing about those changes in the pupil which constitute his school education goes unquestioned. He stands out above all apparatus, all system, all organization. And yet the kind and amount of that apparatus, the nature of that system and organization have a very direct and constant influence upon those subtle and delicate interactions of teacher, pupil, and environment. The character of the organization of the schools has been outlined in the preceding chapter. It remains now to compare the teachers of the different types of schools with respect to some of their qualifications. To hold a position on the teaching staff of the elementary schools of the city of New York two conditions have existed for many years, viz., the possession of the equivalent of a high school diploma and the successful completion of a two-years' teacher training course. It is necessary at present also to take the examination of the Board of Examiners of the City Board of Education. The majority of the teachers receive their professional training at the New York Training School for Teachers and at the Brooklyn Training School for Teachers. A number of high school graduates, however, attend a Liberal Arts College, such as Hunter College. Commencing in the junior year of their attendance at the College, they elect to take the courses in education. These courses along with the experience they receive in practice teaching for which provision is made are accepted in lieu of the two years spent at a training school. At the same time the work in the college is leading to a degree, whereas in the training school such is not the case. It might be said that of the schools in this investigation situated in 20 The Teachers and Pupils 21 Brooklyn most of the teachers are graduates of the Brooklyn Training School for Teachers. Of those schools situated in the Bronx the majority of the teachers took their professional training at the New York Training School for Teachers, but a considerable number are graduates of colleges. In Table V are given the data with respect to the total education after the elementary school of the teachers in the different types of schools. In Table VI are given the types of their pre-service professional training. TABLE V YEARS OF ACADEMIC AND PROFESSIONAL EDUCATION OF THE TEACHERS Percentages of Teachers in Schools with Years of Education after Elementary School School Five Years Six Years Eight Years Nine Years TS................. 91.8 4.1 4.1 D S.................. 18.2 45.5 36.3 R S.................. 14.8 74.1 11.1 TABLE VI PRE-SERVICE PROFESSIONAL TRAINING OF THE TEACHERS IN THE SCHOOLS Percentages of Teachers Taking Professional Training in Normal Schools and Colleges Normal School College School 5 Years after 4 Years after 1 Year 2 Years Elementary High School School T S.................. 91.8 8.2 D S.................. 45.5 18.2 36.3 RS.................. 14.8 74.1 11.1.~~~~~~~~~~~~~~~~~~~~~~~~~~~ From Table V it can be seen that the total number of years of education after the elementary school of the teachers in the experimental schools is slightly greater than that of those in the control schools. But the data in Table VI give evidence of the approximate equality in pre-service professional training of the teachers in Experimental School DS and the control schools with a slight superiority of the teachers in Experimental School TS in this respect. Those teachers who are reported as having attended normal schools only one year, did so before the two-year standard had been adopted. In addition to the observations made earlier in this chapter, 22 Part-Time Types of Elementary Schools TABLE VII COURSES TAKEN IN THE IN-SERVICE PROFESSIONAL TRAINING OF THE TEACHERS WITH NUMBER IN EACH TYPE OF SCHOOL TAKING THEM Schools Courses -- TS DS T RS Total TS DS |R S | Total Methods in Reading.......... 1 Dramatics.....................2 2 Harmony-Music...............1 1 Music-Sight Singing.......... 1 Theory of Music................1 1 Visualization...................1 1 A rt.............................2 2 Class Management.............. 1 7 8 Primary Methods............... 1 1 Program Making................ 1 1 Kindergarten Games............ 1 1 General Methods............... 3 5 3 11 Philosophy..................... 1 1 American Literature............ 1 1 Methods in Arithmetic.......... 1 2 1 4 Summer Session................. 1 1 2 Principles of Education......... 1 1 2 Methods in Drawing.......... 1 1 2 Economics...................... 1 1 Penmanship.................... 2 1 4 Physical Training............... 1 1 1 3 Gregg Shorthand................ 3 3 Vegetable Gardening............ 1 1 Typewriting and Stenography.... 2 1 3 Spanish........................ 1 1 French......................... 2 2 4 International Law and Relations.. 1 English......................... 2 9 11 History......................... 1 1 5 7 Methods in English............. 3 3 6 History of Education........... 1 1 Psychology..................... 1 1 1 3 Principles of Railroad Transportation.....................1 1 Principles of Ocean Transportation 1 1 Physical Education............. 1 Traffic in the Far East........... 1 1 Continuation School Course....1 1 Public School Music.............1 1 2 Bookkeeping and Accountancy..1 1 Drawing and Woodwork...1 1 Education and Exceptional Children........................1 1 Experimental Psychology........1 1 Methods of Teaching History...2 2 Composition and Rhetoric....1 1 Home Economics and Hygiene 1 1 Geography..................2 3 5 Sewing....................1 1 Total. 38 27 48 113~~~~~l Total....................... 38 27 48 113 The Teachers and Pupils 23 with respect to the slight superiority in total years of education of the teachers in the experimental schools over those in the control schools and the longer pre-service professional training of the teachers in Experimental School TS, it may be noted that not only have the teachers of the control schools a longer teaching experience but they have enriched that experience with a greater number of in-service professional courses, so that on the whole there is no evidence of any significant superiority of the teachers in any one of the types of schools. In Table VII are specified the courses which the teachers reported as having taken since they completed their pre-service training. All the courses are given whether they bear directly on the work of the elementary classroom or not. There are very few which will not be at least of indirect value to them. The figures in Table VII give the distribution of the number of times the respective courses were taken by some teacher in the different schools. These figures show that the teachers of the control schools have taken the greatest number of courses, followed by Experimental Schools TS and DS, in turn. Making a rough approximation of the average number of semester hours spent by the teachers in the schools on these courses the figures are as follows: T S...........................................61 D S...........................................102 R S...........................................110 The number of teachers in each of the schools taking some one or more of these courses gives the greatest percentage to the control schools. Of the teachers in the latter schools 75 per cent reported having taken some courses during the years of their teaching, while in the experimental schools TS ancT DS the percentages were 62 per cent and 50 per cent, respectively. EXPERIENCE OF THE TEACHERS The data with respect to the experience of the teachers are analyzed into the facts given in Table VIII. It is evident from Table VIII that the average years' experience of the teachers in the control schools exceeds that of the teachers in the experimental schools. This is true both 24 Part-Time Types of Elementary Schools TABLE VIII TOTAL EXPERIENCE OF THE TEACHERS Percentages of Teachers with Various Lengths of Experience 3 Years 3 to 5 6 to 10 11 to 15 16 to 20 Over 20 Average School or Less Years Years Years Years Years Teer Teacher TS............ 8.3 37.4 33.3 12.5 4.1 4.1 8.6 DS............ 22.8 13.6 31.8 9.1 9.1 13.6 9.4 RS............ 11.1 25.9 22.2 7.4 22.2 11.1 11.2 with respect to total experience and to the number of years spent in their present school. The figures for the latter are as follows: TS, 7.2 years; DS, 5.2 years; and RS, 8.5 years. More data would be required to argue that the holding power of the control type of school is greater than that of the other types, but such data as we have point in that direction. ATTITUDE OF THE TEACHERS Evidence supporting the last statement is to be found in the general preference which teachers throughout the city have for the control type of school. In order to secure a definite statement from the teachers themselves in the two experimental schools with respect to their attitude toward the system in which they were teaching, a questionnaire was prepared for distribution among them. Certain circumstances prevented its distribution among the teachers of Experimental School DS, but from reliable sources the writer received the information that at least 90 per cent of the teachers in this type of school held it in disfavor. They disliked the auditorium periods, but their strongest objection was the daily playground period of fortyfive minutes during which they must keep supervision over their classes. The questionnaire was distributed, however, among the teachers of Experimental School TS. In interpreting the replies to the questionnaire of the teachers in this school, it should be borne in mind that the half-day session with its consequent free morning or afternoon for the teacher is sure to find favor with some. Replies were received from the twenty-four teachers in this school to whom it was sent. Of these, thirteen stated that they did not like the half-day session as well as the regu The Teachers and Pupils 25 lar session, while eleven favored it. Of the thirteen disliking it, eleven expressed a decided preference for the regular session. To the question as to whether they thought they would be able to do as efficient work hour for hour spent in the classroom as they would if the school were on regular session, 11 replied that they were not able to, while 13 thought they could. The opinion of the teachers, however, was less divided on the third question. This was as follows: Do you feel that the part-time type of school, by causing inconveniences with respect to earlier rising and meal hours, or for any other cause, creates an unfavorable attitude in the parents which the children carry over into school and which thus works adversely to their progress in school? To this question eight gave an affirmative reply while sixteen thought it had no effect. In fine, it may be noted that the physical and moral effects of the respective systems upon the boys and girls is another aspect of the whole problem, very important, but lying outside the purview of this study. But, so far as the effect of the parttime system upon the work of the teachers is concerned, the statement might safely be made that any feeling of disfavor toward the system in which the teacher is working is not entertained in such a personal way that he does not give of his best endeavor to the work in which he is daily engaged. 2. THE PUPILS BIRTHPLACE OF PARENTS The parents of the pupils attending the four schools selected for this investigation are for the most part of foreign origin. In some, the process of Americanization has proceeded farther than in others. For instance, though the majority of the parents in the community of Control School RS are of foreign origin, yet most of these were born in this country. In Table IX are given the percentages of the total number reporting whose parents were born in the United States or in other countries. From Table IX it is evident that in Experimental School TS there is the highest percentage of pupils with parents of foreign birth. It is not to be inferred from this, however, that the language problem in this school in the grades selected for the investigation is an acute one. Further data gathered showed 26 Part-Time Types of Elementary Schools TABLE IX BIRTHPLACE OF THE PARENTS OF THE PUPILS OF THE SCHOOLS Total Number of Pupils Reporting and Percentages of Parents Born in the United States * and in Foreign Countries Number of Both Parents One Parent Neither Parent School Pupils Born in the Born in the Born in the Reporting United States United States United States TS................. 621 13.5 19.4 67.1 DS................. 621 33.3 12.4 54.3 RS................. 503 64.4 12.1 23.5 * Under the United States are counted all those parents born in Englishspeaking countries. that approximately one-third of these parents received their education in whole or in part in this country. And yet it is obvious that the first generation of foreign immigration is more numerous in the communities of the experimental schools than in those of the control schools, and whatever handicap this fact involves in pupil achievement belongs to the former schools. THE AGE OF THE PUPILS In a study of this nature the facts concerning the ages of the pupils are usually made the basis of some conclusions regarding the degree of acceleration and retardation made by these pupils. In this investigation to make such a study relevant and its conclusions valuable as bearing upon the effects' of the several types of organization upon the problem of acceleration and retardation would have demanded reliable information as to the type of school which each pupil had attended during his whole school career. This was impossible for two very closely related reasons. First, for a number of years previous to 1922, when this investigation was made, the school population was of a very shifting character; and, second, owing to this fact, it was impossible to secure accurate data in connection with a sufficiently large number of pupils who had attended the same system during all their school life to make any conclusions based on this study at all cogent. However, it will throw considerable light upon the relative ages of the pupils in the three types of schools if the pupils are distributed into groups according to their being of normal age, under-age, or over-age. The same standards will be adopted The Teachers and Pupils 27 for this study as were used by Nifenecker 1 in a study on pupils' progress through the grades on the basis of the register of all the schools of the City of New York on February 28, 1921. For normal progress these were as follows: Grade 1, 6 to 7Y2 years; Grade 2, 7 to 82 years, etc. It might be stated that in the study just referred to 12.09 per cent were reported as having made rapid progress; 41.9 per cent, normal progress; and 45.9 per cent, slow progress. In Table X are given the percentages for these three divisions of the pupils who were included in the present study. TABLE X PERCENTAGE OF PUPILS OF NORMAL AGE, UNDER-AGE, AND OVER-AGE IN THE THREE TYPES OF SCHOOLS School Under-Age Normal Age Over-Age TS..................... 31.3 44.6 24.1 D S..................... 27.5 41.1 31.4 RS..................... 36.8 39.8 23.4 The most significant fact to notice from Table X is the low percentage of over-age pupils in all the schools in comparison with the figures for the city as a whole the previous year. It is also evident that Experimental School DS has the highest percentage of over-age pupils and the control schools the highest percentage of under-age pupils. This same fact is indicated by the average age of the pupils in each of the schools. The averages are as follows: Experimental School TS, 10 years, 8 months; Experimental School DS, 10 years, 91/ months; Control Schools, 10 years, 6 months. Other things being equal, to be over-age is an advantage, but this statement should be considered in the light of scores on the intelligence test which will show that the lower chronological age of the pupils of the control schools has not given them an average lower score on the intelligence test. ATTENDANCE Bachman 2 found in a study of the causes of non-promotion in the New York City Schools that regularity of attendance 1 Nifenecker, Eugene A. Pupils' Progress through the Grades. Board of Education, City of New York, Bureau of Reference, Research, and Statistics, New York, 1922, pp. 24, 53. 2 Bachman, Frank P. Problems in Elementary School Administration. World Book Company, Yonkers-on-Hudson, 1915, p. 138. 28 Part-Time Types of Elementary Schools was an important factor. Of those absent 20 days and less, more than 90 per cent secured promotion; of those absent 21 days and more, the percentage of promotions was less than 75. It is probable that the effect of absence is not significant until such absence is large in amount. Data on the attendance of the pupils in this study were secured for the two full terms immediately preceding the one in which the testing program was carried on, as also for the two months of that term. The total possible attendance was thus 233 days. In Table XI this total time is divided into four parts and the percentages of pupils in each type of school whose attendance fell between these divisions are given. TABLE XI ATTENDANCE OF PUPILS DURING 233 SCHOOL DAYS IN THE THREE TYPES OF SCHOOLS Percentage of Pupils with Attendance Falling in Each Division 225 to 233 215 to 224 200 to 214 Less Than School Days Days Days 200 Days T S.................. 3.8 41.0 35.4 19.8 D S............. 18.4 40.9 26.7 14.0 R............. 31.6 36.2 21.7 10.5 The figures of Table XI show that in regularity of attendance the control schools have a distinct superiority. If the pupils were divided into two classes for the 233 days on the basis of the division used by Bachman above, viz., an absence of 20 days and less, and of 21 days and more, during one term, the same superiority for the control schools would be shown. The percentages of the pupils attending at the rate of less than 20 days per term for the schools is as follows: T S.........................................6.1% D S.........................................6.0% R S.........................................3.0% It is doubtful, however, whether the small difference between the experimental schools and the control schools in the number of pupils whose attendance has dropped to the point where it adversely affects their standing would have much effect on the comparative educational status of the school as a whole. CHAPTER V PUPIL ACHIEVEMENT THE TESTS In this investigation the examination of the pupils was confined to the fundamentals of the school curriculum. To measure the differences in the mental ability of the groups an intelligence test was given and in the light of this important variable the results in the educational tests were interpreted. The tests used in the investigation were as follows: Intelligence: National Intelligence Test, Scale A, Form 1 Educational: Reading-Thorndike-McCall Reading Scale, Form 3 and Burgess Scale for Measuring Speed in Silent Reading PS-1 Arithmetic-Woody-McCall Mixed Fundamentals, Form 1 Spelling-Horizontal Section of the Ayres Scale Handwriting-Thorndike Handwriting Scale. As was indicated in an earlier chapter the tests were given in the schools by forty senior students of the New York Training School for Teachers. These students were thoroughly competent and did this part of the work efficiently. To insure further uniformity in the testing program a request was made and willingly acceded to that the regular classroom teacher be absent from the room while the tests were being given. A report was received from each of the forty students for the purpose of recording any unusual or irregular circumstances during their time in the classroom. Where there were any connected with a pupil's test his papers were marked and eliminated from the study. There were very few which were thus eliminated. The tests were given on two days, March 29 and 30. The scoring of all tests was done under the direct supervision of the writer. This secured uniformity throughout this part 29 30 Part-Time Types of Elementary Schools of the work. In addition all the tests were checked by him after scoring. The samples of the pupils' handwriting were scored for quality by three persons, one of whom was the writer, and the average of the three scores was taken as the final evaluation of the quality of the pupil's handwriting. The statistical section of the study was checked throughout. STATISTICAL TREATMENT TO SECURE COMPARABLE MEASURES OF DIFFERENCES An examination of any random number of standard tests reveals the marked lack of uniformity in their range of scores. One may run, for example, from zero to 100; another from zero to 35. It is evident that the differences between the scores 28 and 32 on each of these two tests do not represent equal amounts. In educational measurement, however, it is frequently desired to compare such differences, but in order to be able to do so, it is necessary first to resolve the scores in the different tests into comparable units. One of the phases of statistical measurement which has emphasized the necessity of having such comparable units has been the recognized desirability of using intelligence tests in connection with educational tests for purposes of comparing groups in educational achievement. The amount of mental ability possessed sets very definite limits to the amount of achievement possible. Groups with unequal mental ability should not be expected to achieve equally in the work of the classroom. But with unequal scores made by two groups in both an intelligence test and a reading test before us-tests whose ranges of scores are unequal-the problem at once arises how to determine by a comparison of the scores which of the two groups made the highest achievement in the reading test interpreted in the light of their mental ability. Pintner1 was one of the earliest to recognize the problem and to secure a solution to it. He did this by using the percentile method of scaling both the mental and educational tests. The problem has also been met by converting the scores into an age scale. This gives a mental age and an educational age. These can be directly compared. Out of these comparable units have been evolved the Achievement Pintner, R. The Mental Survey. D. Appleton and Company, New York, 1918, Chap. 6. Pupil Achievement 31 Quotient (AQ) by Buckingham and Monroe and the Accomplishment Ratio (AR) by Franzen. The formula for each is as follows: AQ (AR) Educational Age Mental Age It is generally recognized that one of the best ways of comparing different groups in abilities is based upon their averages and some measure of their variability such as the standard deviation. The standard deviation is doubtless the most satisfactory of all dispersion measures. If it were possible, then, to resolve the scores in all the tests into some such measure, there would be no difficulty in making such comparisons as have been mentioned. The T scale 2 affords such units and has been adopted for use in this study. The main problem, then, is clear, viz., how to convert the scores in all of the tests into these T-scale units. The solution of this problem is based on an assumption upon which all tests are scaled, that is, the average variability of a group of pupils in one test is the same as in another. Consequently, in a real sense there is such a thing as a correspondence 3 between tests. Accepting the assumption that pupils in large groups differ from one another equally in two tests, and that there is thus a resultant correspondence between them, the first step in transmuting the crude scores of the tests in this study into T-scale units was to select a test which had previously been scored in these units. This test was McCall's Multi-Mental Scale. Against this test was calibrated the tests used in the investigation. This was done through the National Intelligence Test whose scores had previously been transmuted into T units by being calibrated against the multi-mental scale. Basing our explanation of the method of calibrating one test against another on the two assumptions made in the preceding paragraph, it will follow that the ability represented by the total range of the scores in one test is equal to that represented by the total range of scores in another test. Hence, the highest and lowest scores in the one test correspond with those of like rank in the second, though these may not be made by the same 'McCall, W. A. How To Measure in Education. Macmillan Co., New York, 1923, Chap. X. 'Otis, A. S. Statistical Method in Educational Measurement. World Book Company, Yonkers, 1925, Chap. X. 32 Part-Time Types of Elementary Schools pupil. Likewise the median score will correspond with the median and any other point will measure with its corresponding measure in the other test. As was just stated the first test thus calibrated was the National Intelligence Test. Thereafter the percentiles of the distributions in the other tests were plotted against the percentile scores of the pupils in this. The second, ninety-eighth and every fifth percentile were found. On the horizontal line of the graph were placed the scores in the National Intelligence Test in T units, and on the vertical line the crude scores of the test being calibrated. Each of the percentile scores of thie latter test was plotted against its equivalent score of the National Intelligence Test which was now found in T-scale units. A line was drawn which passed most closely through the points of intersection thus formed. This line became the line of relation between the two tests. From this line of relation could be read the equivalent T scores for the crude scores of the second test. All the tests were calibrated against the Multi-Mental Scale through the National Intelligence Test rather than through each successive test to prevent errors from being cumulative. THE F TECHNIQUE Reference has already been made in the preceding section to the value of an intelligence test in comparing the achievement of groups as measured by educational tests. Further reference was made to the fact that in the AQ and the AR techniques the scores in the intelligence and educational tests are transmuted into age-scale units in order that comparisons might be made between them. The formula for these techniques was also given. The advantages gained by these methods were secured for the T scale by the adoption of equivalent T-scale units 4 for the AQ and AR techniques. The formula thus evolved is as follows: F =Te-Ti+50=Effort or efficiency Fr-=Tr-Ti+50=Effort in reading Fa=Ta-Ti-50 —Effort in arithmetic McCall explains the formula in these words: "Ti is merely a T score on some intelligence test. Te is the average T score on several educational tests. Tr is the T score on some reading 'McCall, W. A. How To Ewperiment in Education. Macmillan Company, New York, 1923, p. 103. Pupil Achievement 33 test. Ta is the T score on some arithmetic test. Each F is explained by its formula." Where the pupils reach a standard of educational achievement equal to their mental capacity to achieve, the figures to the right of the equation will be 50. When this score is less, the standard for some reason has not been reached; where above 50, the standard has been surpassed. No separate tables have been given of the scores of the pupils in the intelligence and educational tests. In the Tables which follow, making use of the F formula explained above, these scores for the grades in each type of school may be studied. Table XII should be read as follows, and each of Tables XIII to XVII similarly: The F score for Grade 3B in Experimental School TS in the Thorndike-McCall Reading Scale is 50.95. This is found by adding 50 to the difference between the average, or mean T score, made by the pupils in Grade 3B in this test and the National Intelligence Test. TABLE XII F SCORES BY GRADES IN THE THORNDIKE-MCCALL READING THREE TYPES OF SCHOOLS SCALE FOR THE Mean Mean F Grade School Reading N.I.T. Score T Score T Score TS.......................35.81 - 34.86 + 50 - 50.95 3B DS.......................34.83 - 35.36 + 50 49.47 RS....................... 32.03 - 33.60 + 50 - 48.43 TS.......................37.14 - 37.65 + 50 = 49.49 4A DS.......................39.79 - 39.07 + 50 50.72 RS.......................36.69 - 37.13 + 50 - 49.56 TS..................... 40.79 - 42.48 + 50 = 48.31 4B DS.......................41.09 - 41.66 + 50 49.43 RS.......................40.03 - 41.95 + 50 = 48.08 TS.......................42.31 - 43.51 + 50 48.80 5A DS......................43.55 - 45.16 + 50 48.39 RS.......................41.55 - 45.84 + 50 45.71 TS.......................47.81 - 48.73 + 50 = 49.08 5B DS.......................45.97 - 46.87 + 50 = 49.10 RS.......................45.37 - 46.46 + 50 = 48.91 TS.......................46.74 - 47.71 + 50 = 49.03 6A DS.......................49.03 - 48.57 + 50 = 50.46 RS.......................47.24 - 49.10 + 50 48.14 TS.......................50.47 - 50.20 + 50 = 50.27 6B DS.......................51.02 - 49.21 + 50 51.81 RS.....................48.31 - 52.42 + 50 = 45.89 34 Part-Time Types of Elementary Schools TABLE XIII F SCORES BY GRADES IN THE BURGESS SCALE FOR MEASURING SPEED IN SILENT READING FOR THE THREE TYPES OF SCHOOLS Mean Mean F Grade School Reading N.I.T. Score T Score T Score TS.......................36.30 - 34.86 + 50 = 51.44 3B DS.......................33.70 - 35.36 + 50 = 48.34 RS....................... 34.30 - 33.60 + 50 = 50.70 TS.......................37.10 - 37.65 + 50 - 49.45 4A DS...................... 39.50 -- 39.07 + 50 = 50.43 RS.................. 36.40 - 37.13 + 50 = 49.27 TS.....................41.50 - 42.48 + 50 = 49.02 4B DS 42.70 - 41.66 + 50 = 51.04 RS.....................40.10 - 41.95 + 50 = 48.15 TS.......................42.40 - 43.51 + 50 = 48.89 5A DS.....................43.70 - 45.16 + 50 = 48.54 RS.......................43.70 - 45.84 + 50 = 47.84 TS.....................46.40 - 48.73 + 50 = 47.67 5B DS.......................46.60 - 46.87 + 50 = 49.73 RS....................... 43.80 - 46.46 + 50 = 47.34 TS.......................46.70 - 47.71 + 50 = 48.99 6A DS.......................45.10 48.57 + 50 = 46.53 6 A D S.45.10 - 48.57 + 50 = 46.53 RS....................... 46.50 - 49.10 + 50 = 47.40 TS.......................48.10 - 50.20 + 50 = 47.90 6B DS.......................47.40 - 49.21 + 50 = 48.19 RS. v....................47.50 - 52.42 + 50 = 45.08 From an examination of the test scores given in Tables XIIXVIII the following facts may be noticed. 1. Though the difference is very small, the average T score in the National Intelligence Test for all the pupils in each of the three types of schools is seen from Table XVIII to be highest in the control schools and lowest in the Experimental School DS. These higher scores are evident also in three of the seven grades when reference is made to the intelligence scores in any of the other tables. 2. The efficiency of the pupils in the control schools in both the Reading Tests is lower than that of the pupils in the experimental schools. Of the latter schools, the pupils in Experimental School DS are superior. 3. The F scores made by Experimental School DS in arithmetic were the highest made by any of the three types of schools Pupil Achievement 35 TABLE XIV F SCORES BY GRADES IN THE WOODY-MCCALL MIXED FUNDAMENTALS IN ARITHMETIC FOR THE THREE TYPES OF SCHOOLS Mean Mean Grade School Arith- N.I.T. F Grrade School metic T Score T Score Score TS.......................31.40 - 34.86 + 50 - 46.54 3B DS.......................34.10 - 35.36 + 50 = 48.74 RS.......................31.10 - 33.60 + 50 = 47.50 TS.......................34.10 - 37.65 + 50 = 46.45 4A DS.......................39.50 - 39.07 + 50 50.43 RS.......................35.10 - 37.13 + 50 - 47.97 TS....................... 38.80 - 42.48 + 50 = 46.32 4B DS.......................42.30 - 41.66 + 50 = 50.64 RS...................... 40.50 - 41.95 + 50 = 48.55 TS.......................42.00 - 43.51 + 50 = 48.49 5A DS.......................44.40 - 45.16 + 50 49.24 RS.......................42.50 - 45.84 + 50 == 46.66 TS.......................47.70 - 48.73 + 50 - 48.97 5B DS.......................49.60 - 46.87 + 50 = 52.73 RS.......................49.50 - 46.46 + 50 = 53.04 TS.......................49.10 - 47.71 + 50 - 51.39 6A DS.......................54.20 - 48.57 + 50 - 55.63 RS.................... 49.00 - 49.10 + 50 = 49.90 TS.......................54.90 - 50.20 + 50 = 54.70 6B DS.......................53.50 - 49.21 + 50 - 54.29 RS.......................53.50 - 52.42 + 50 = 51.08 in any of the tests. The work done in this school in this subject is well above average. The results in this test were slightly lower in Experimental School TS than in the control schools. 4. The efficiency of the pupils in reading as measured by the F scores is superior on the whole for the lower grades in all the schools than for the higher. The reverse of this is true in arithmetic. 5. In spelling, the F scores of the pupils in Experimental School TS were highest while those of the control schools were lowest. 6. In quality of handwriting Experimental School DS secured the highest standing, but in speed Experimental School TS ranked first. 7. Combining all the subjects and grades, thus giving the total educational achievement for the school as a whole, Table 36 Part-Time Types of Elementary Schools TABLE XV F SCORES BY GRADES IN THE SPELLING TEST FOR THE THREE TYPES OF SCHOOLS Mean Mean F Grade School Spelling N.I.T. Score T Score T Score TS....................... 34.84 - 34.86 + 50 = 49.98 3B DS.......................32.19 - 35.36 + 50 = 46.83 RS.......................34.51 - 33.60 + 50 = 50.91 TS.......................39.61 - 37.65 + 50 = 51.96 4A DS.....................37.04 - 39.07 + 50 = 47.97 RS.......................39.30 - 37.13 + 50 = 52.17 TS.......................43.40 - 42.48 + 50 = 50.92 4B DS.....................40.00 - 41.66 + 50 = 48.34 RS.......................39.85 - 41.95 + 50 = 47.90 TS.......................45.36 - 43.51 + 50 = 51.85 5A DS.......................41.56 - 45.16 + 50 = 46.40 RS.......................44.29 - 45.84 + 50 = 48.45 TS.......................49.02 - 48.73 + 50 = 50.29 5B DS.......................45.84 - 46.87 + 50 = 48.97 RS......................41.80 - 46.46 + 50 = 45.34 TS...................... 48.21 - 47.71 + 50 = 50.50 6A DS.......................45.99 - 48.57 + 50 = 47.42 RS.......................48.30 - 49.10 + 50 = 49.20 TS.......................51.83 - 50.20 + 50 = 51.63 6B DS.......................50.32 - 49.21 + 50 = 51.11 RS.......................49.65 - 52.42 + 50 = 47.23 XVIII-B shows a small but real difference between the schools in favor of those of the experimental types with Experimental School DS ranking highest. 8. Further support of this general finding lies in the fact that out of 42 rankings in the 6 tests through the seven grades, the control schools ranked highest in only 6 instances and lowest 22 times. The two types of experimental schools each ranked highest in 18 cases. EQUIVALENT GROUPS METHOD 5 The method used in the preceding section in the comparison of the achievement of the pupils in the three types of schools took into account only one variable, namely, intelligence. It will be granted that this is a very important one, but there were a McCall, W. A. How To Experiment in Education. Macmillan Company, New York, 1923, Chap. VII. Pupil Achievement 37 TABLE XVI F SCORES BY GRADES IN THE HANDWRITING (QUALITY) TEST FOR THE THREE TYPES OF SCHOOLS Mean Hand- Mean Grade School writing N.I.T. F (Quality) T Score T Score Score TS.......................35.99 - 34.86 + 50 = 51.13 3B DS......................39.65 - 35.36 + 50 = 5429 RS.......................33.94 - 33.60 + 50 = 50.34 TS.......................37.95 - 37.65 + 50 = 50.30 4A DS.......................40.80 - 39.07 + 50 = 51.73 RS....................... 39.79 - 37.13 + 50 = 52.66 TS................. 37.35 - 42.48 + 50 = 44.87 4B DS............... 42.23 - 41.66 + 50 = 50.57 RS.......................42.65 - 41.95 + 50 = 50.70 TS.......................42.68 - 43.51 + 50 = 49.17 5A DS.......................42.07 - 45.16 + 50 = 46.91 RS....................... 44.95 - 45.84 + 50 = 49.11 TS.......................39.44 - 48.73 + 50 = 40.71 5B DS.......................46.50 - 46.87 + 50 = 49.63 RS.......................44.54 - 46.46 + 50 = 48.08 TS.......................43.90 - 47.71 + 50 = 46.19 6A DS.......................43.78 - 48.57 + 50 = 45.21 RS.......................43.16 - 49.10 + 50 = 44.06 TS.......................44.72 - 50.20 + 50 = 44.52 6B DS.......................47.45 - 49.21 + 50 = 48.24 RS......................45.84 - 52.42 + 50 = 43.42 others, however, such as age, regularity of attendance, and birthplace of parents which might be felt to have influenced the results recorded there. In order to take account of these and other variables, the equivalent groups method was also utilized in the treatment of the data. The procedure by which the equivalent groups from the different types of schools were secured will now be explained. The names of all the pupils from each type of school were written down in the order of the size of their scores on the National Intelligence Test. Then were written opposite their respective names the data with respect to the following variables: sex, chronological age, birthplace of parents, grade, number of terms at school, days' attendance and the intelligence score. Then an attempt was made to pair the pupils with respect to these seven variables. Two schools only were taken at a time. A 38 Part-Time Types of Elementary Schools TABLE XVII F SCORES BY GRADES IN THE HANDWRITING (SPED) TEST FOR THE THREE TYPES OF SCHOOLS Mean Hand- Mean F Grade School writing N.I.T. Score (Speed) T T Score Score TS.......................35.53 - 34.86 + 50 = 50.67 3B DS.......................35.12 - 35.36 + 50 = 49.76 RS.......................43.15 - 33.60 + 50 = 59.55 TS.......................40.54 - 37.65 + 50 = 52.89 4A DS.......................37.02 - 39.07 + 50 = 47.95 RS.....................37.44 - 37.13 + 50 = 50.31 TS....................... 41.93 - 42.48 + 50 = 49.45 4B DS.......................38.34 - 41.66 + 50 = 46.68 RS.......................41.05 - 41.75 + 50 = 49.30 TS......................45.03 - 43.51 + 50 = 51.52 5A DS.......................43.59 - 45.16 + 50 = 48.33 RS.......................44.57 - 45.84 + 50 = 48.73 TS.......................49.01 - 48.73 + 50 = 50.28 5B DS.......................44.58 - 46.87 + 50 = 47.71 RS.......................44.36 - 46.46 + 50 = 47.90 TS.....................48.22 - 47.71 + 50 = 50.51 6A DS.......................49.20 - 48.57 + 50 = 50.63 RS.......................48.97 - 49.10 + 50 = 49.87 TS.......................49.73 - 50.20 + 50 = 49.53 6B DS.......................46.35 - 49.21 + 50 = 47.14 RS.......................49.52 - 52.42 + 50 = 47.10 pupil was taken from one school and an effort made to pair this pupil with one from the other school. It was impossible to find a child in school DS, for instance, who could be paired with one in school RS with all these variables exactly equated. It was decided then to keep constant sex, number of terms at school, and birthplace of parents, and to let chronological age vary not more than five months, grade not more than one, attendance not more than an average of eight days per term, and the National Intelligence score not more than five points in crude scores. Care was taken to keep the differences between the schools in any one variable as near zero as possible. As a result of this, the average age, days' attendance, etc., for each group were practically equal. For example, there were 111 pairs equated between schools DS and RS. The average differences in the total days' attendance between the two groups was.6 of a Pupil Achievement 39 TABLE XVIII SUMMARY OF TABLES XII-XVII A-F SCORES FOR ALL TESTS WITH ALL GRADES TOGETHER B-F SCORES WITH ALL GRADES AND ALL SUBJECTS TOGETHER FOR THE THREE TYPES OF SCHOOLS Mean Mean Educa- N.I.T. F Test School tional T Score T Score Score A Thorndike-McCall TS.............. 43.00 - 43.61 + 50 49.39 Reading DS............. 43.47 - 43.54 + 50 = 49.93 RS.............41.73 - 43.87 + 50 = 47.86 Burgess Silent TS............. 42.66 - 43.61 + 50 = 49.05 Reading DS.............. 42.53 - 43.54 + 50 = 48.99 RS............. 41.77 - 43.87 + 50 = 47.90 Woody-McCall TS.............. 42.52 - 43.61 + 50 = 48.91 Arithmetic DS.............. 45.22 - 43.54 + 50 = 51.68 RS.............43.26 - 43.87 + 50 = 49.39 Spelling TS..............44.63 - 43.61 + 50 = 51.02 DS............ 42.34 - 43.54 + 50 = 48.80 RS............42.00 - 43.87 + 50 = 48.13 Handwriting TS..............40.25 - 43.61 + 50 = 46.64 Quality DS........... 43.18 - 43.54 + 50 = 49.64 RS.............42.19 - 43.87 + 50 = 48.32 Handwriting TS.............. 44.30 - 43.61 + 50 = 50.79 Speed DS..............41.87 - 43.54 + 50 = 48.33 RS.............44.12 - 43.87 + 50 = 50.25 B All Tests TS..............42.89 - 43.61 + 50 - 49.28 DS............ 43.10 - 43.54 + 50 = 49.56 RS............ 42.51 - 43.87 + 50 - 48.64 day in favor of School RS; in the National Intelligence Test an average score of.1 in favor of the same school; and in age an average of.2 month in favor of School DS. These differences are negligible. It might be pointed out further that "birthplace of parents" should be interpreted with respect to whether the parents were born in an English-speaking country or not. Pupils, both of whose parents were from non-English-speaking countries, were paired, as also were those with both parents from Englishspeaking countries. Where one parent was foreign and the other not, the child was paired either with a child of like parentage or 40 Part-Time Types of Elementary Schools with one, both of whose parents were born in an English-speaking country. Proceeding in the foregoing way, 111 pairs were equated in Schools DS and RS; 102 in Schools TS and RS; and 121 in the two experimental schools. The data in connection with the equated groups are on file in the Library of Teachers College, Columbia University. After the pairs were thus equated, the scores made by each of the pupils in the several tests were written opposite their names. Then these scores were transmuted into T-scale units, and the mean T score for each of the equated groups found in each of the tests. The differences between these means were determined and the P.E. of these differences found. This last measure gives the reliability of the differences in the means between the two groups. Table XIX shows the mean T scores in the tests for the equated groups. The Table should be read as follows: The mean T score in the Thorndike-McCall Reading Scale is 43.833 for Experimental School TS; the T score for the control schools is 41.069; the difference is 2.764 and the P.E. of the difference is ~.740. These figures mean that there is found a difference between the two types of schools of 2.764 T units in comprehension in reading as measured by the Thorndike-McCall Reading Scale, and that the measure of the reliability of this difference is plus and minus the P.E. The remainder of the table should be read in like manner. The significance of this difference of 2.764 between the groups in the two types of schools may be judged by the following: 6 The chances are 1 to 1 that the true measure is between obtained measure i 1 P.E. measure. The chances are 4.6 to 1 that the true measure is between obtained measure ~ 2 P.E. measure. The chances are 22 to 1 that the true measure is between obtained measure ~ 3 P.E. measure. The chances are 142 to 1 that the true measure is between obtained measure ~ 4 P.E. measure. The chances are 369 to 1 (practical certainty) that the true measure is between obtained measure ~+ 4.4 P.E. measure. *McCall, W. A. How To Measure in Education. Macmillan Company, New York, 1922, p. 406. Pupil Achievement 41 If the difference of the means, then, is 4.4 X P.E., the difference may with certainty be considered a real and significant one. This degree of reliability will be used in the interpretation of the results in this study. It will be noted in Table XIX that the mean T score of the group from Experimental School TS when taken with Control School RS is not the same as the mean when taken with Experimental School DS. This is due to the fact that the pupils who were selected from School TS when being paired with pupils from School RS were not the same as those who were selected when the pupils of School TS were being paired with those from School RS. TABLE XIX MEAN T SCORES IN THE TESTS FOR THE EQUATED GROUPS FROM THE THREE TYPES OF SCHOOLS, THE DIFFERENCES (D) B'EWEEN THE MEANS AND THE PROBABLE ERROR (P. E.) OF THE DIFFERENCES Test TS DS RS D P.E. Thorndike-McCall Read- 43,833 41.069 2.764.740 ing.................. 42.198 40.288 1.910.616 43.325 43.375 -.050.675 Burgess Reading....... 42.314 41.441.873.724 41.360 40.081 1.279.635 43.375 43.125.250.657 Woody-McCall Arithme- 43.206 43.108.098.933 tic.................. 43.099 41.279 1.820.745 44.142 44.908 -.766.825 Spelling................ 44.598 42.127 2.471.693 41.766 40.351 1.415.581 44.750 43.450 1.300.621 Writing-Quality....... 40.373 41.431 -1.058.633 42.604 41.495 1.109.588 40.358 43.675 -3.317.527 Writing-Speed........ 686 43.127.559.658 40.883 43.045 -2.162.640 44.542 40.983 3.559.764 An examination of Table XIX justifies the following comments in connection therewith: 1. The differences of 3.7 X P.E. and 3.1 X P.E. in the Thorndike-McCall Reading Scale between the Experimental Schools TS and DS, respectively, and the control schools, and of 3.6. X P.E. in spelling between Experimental School TS and the control schools, though not reaching the limit generally 42 Part-Time Types of Elementary Schools accepted in determining the reliability of differences, challenge attention and tend to support the findings from the study of the test results by the F technique. 2. Considering quality and speed as two phases of the one function of handwriting, the differences between the schools are very small. Experimental School DS excels Experimental School TS by a real and significant difference in quality; but with respect to speed, the reverse is true. In the other tests the differences between these schools are small and insignificant. 3. With the one exception of handwriting, all the differences in the test results between the experimental and control types of schools are in one direction, viz., in favor of the experimental schools. THE LOWER AND UPPER HALVES OF THE EQUATED GROUPS In addition to finding the differences between the groups as was done in the preceding section where they were taken as wholes, it was desired to discover if there were any significant differences between the schools when the groups were compared in two parts, viz., the lower with the lower and the higher with the higher. The lower and higher sections of the groups will be understood if it is recalled how the groups were formed. The first step was to list the pupils irrespective of age or grade or any other variable in groups with crude scores on the National Intelligence Test within a range of five points, starting with the lowest scores and proceeding to the higher. This caused, then, the lower half of the group to have the lower crude scores in the National Intelligence Test and the upper half, the higher scores. The place of division into halves is approximately where the crude scores on the intelligence test were 88. It happened also that for the most part the younger pupils and those in the lower grades were in the lower halves of the groups. In this comparison of the lower and upper halves, respectively, the handwriting test was omitted as was also the comparison of the two experimental schools between which in the fundamental subjects no differences approaching significance were found in the previous study as reported in Table XIX. The explanation given in the preceding section with respect to the reliability of the differences between the means of the two equated groups should be applied also in the interpretation of Pupil Achievement 43 the results found in Table XX. Table XX should be read as follows: The mean T score in the Thorndike-McCall Reading Scale for the lower half of the equated group in the Experimental School TS is 39.275, the mean T score for the lower (L) half in the control schools is 36.196, the difference is 3.079 and the P.E. of the difference is ~.741. This is followed by the difference in the mean T scores of the upper (U) halves. TABLE XX MEAN T SCORES IN THE TESTS FOR THE UPPER AND LOWER HALVES OF THE EQUATED GROUPS FROM THE THREE TYPES OF SCHOOLS, THE DIFFERENCES (D) BETWEEN THE MEANS AND THE PROBABLE ERROR (P. E.) OF THE DIFFERENCES Test Half TS DS R S D P.E. Thorndike- L............ 39.275 36.196 3.079.741 McCall Reading U............ 48.392 45.941 2.451.920 L............ 38.545 37.182 1.363.740 U............ 45.786 43.339 2.447.777 Burgess L............ 38.039 37.412.627.861 Reading U............ 46.588 45.471 1.117.853 L........... 36.964 36.945.019.699 U........... 45.679 43.161 2.518.808 Woody-McCall L............ 37.373 37.725 -.352 1.040 Arithmetic U............ 49.039 48.490.549 1.126 L........... 39.891 36.273 3.618.874 U............ 46.250 46.196.054.943 Spelling L............ 40.039 37.471 2.568.818 U............ 49.157 46.784 2.373.701 L............ 38.218 36.818 1.400.696 U............ 45.250 43.821 1.429.683 1. The superiority of Experimental School TS over the control schools in the Thorndike-McCall test for comprehension in reading which was noted in the discussion of the previous table is seen in Table XX to exist particularly in the lower halves of the groups. Here the difference 4.2. P.E. approaches closely to being a real and significant one. It is striking that there is practically no difference between the lower sections of these groups in the Burgess test for speed in reading. 2. With the exception of the instance just mentioned, the experimental schools show a greater superiority over the control schools in reading in the upper halves of the groups than in the lower. 44 Part-Time Types of Elementary Schools 3. The difference 4.1 P.E. between the lower halves of the groups of Experimental School DS and the control schools in the Woody-McCall test in arithmetic is almost as real and significant as the one in comprehension in reading mentioned above. 4. The differences in spelling are in favor of the experimental schools but these differences are not significant. CHAPTER VI SUMMARY AND CONCLUSIONS GENERAL SUMMARY 1. The two part-time types of schools selected for this study were the morning and afternoon session, or two-session (TS) type, and the duplicate session (DS) type. The regular session (RS) school was used as the control type. 2. The representatives of the different types of schools were selected with a view to having them as similar as possible in such factors as social status of parents, freedom of the pupils from language difficulties because of non-English homes, size of school, etc. A study of the data with respect to these factors reveals no significant differences among the three types of schools included in this investigation. 3. Altogether 1,786 pupils are included in the study. All these were under the system in which they were tested for two years. This number is divided among the schools as follows: Experimental School TS, 659; Experimental School DS, 621; and the control schools, 506. 4. The average enrollment in the classes is highest in the experimental schools but not sufficiently high to be a factor in retarding progress. 5. The school programs present the outstanding features which differentiate the schools in organization. 6. The amounts and systems of supervision are similar in all the schools. 7. Experimental School TS promotes a larger percentage of classroom enrollment than do the other schools. 8. On the whole the teachers of the experimental schools excel those of the control schools in length of academic and preservice professional training. 9. Offsetting the slightly superior pre-service preparation of the teachers in the experimental schools is the more extensive 45 46 Part-Time Types of Elementary Schools experience of the teachers of the control schools enriched with a greater number of in-service professional courses. 10. Whatever may be the personal attitude of the teacher towards the system of organization in which he is teaching, according to the judgments of the principals this attitude affects but slightly, if at all, his classroom work. 11. There is a larger percentage of pupils whose parents came from foreign countries in the experimental schools than in the control schools. Whatever handicap this fact involves belongs to the former schools. 12. On the whole the pupils of the control schools are younger than those of the experimental schools. This is somewhat more evident when the comparison is made with Experimental School DS. 13. The regularity of attendance in the control schools is distinctly greater than in the experimental schools. 14. The study of pupil achievement in the three types of schools was made by means of a testing program, including tests for intelligence, comprehension and speed in reading, arithmetic, spelling, and handwriting. 15. The comparison of the test results in the schools was made by two methods. First, the crude scores in all the tests were transmuted into T-scale units and the F, or efficiency technique, used to interpret the test results. Second, after equating groups of more than a hundred pupils from each of the schools with respect to a number of important variables, the equivalent-groups method was used whereby the differences between the groups were found and the reliability of those differences determined. 16. Though the average chronological age of the pupils of the control schools is less than that of the experimental schools, they have a higher average score on the National Intelligence Test than have the pupils of the latter schools. 17. Notwithstanding this higher average intelligence score of the pupils in the control schools, the evidence is cumulative throughout the treatment of the test results of a slight superiority of the pupils in the experimental schools over those of the control schools in total educational efficiency. In F scores for all the pupils in all the tests this is indicated as follows: TS, 49.28; DS, 49.56; and RS, 48.64. Summary and Conclusions 47 18. The results from the Equivalent-Groups Method, also, give consistent, though small, differences in favor of the experimental schools, with the one exception of handwriting. 19. The most important differences between the experimental schools and the control schools in reading are in the ThorndikeMcCall test. When the halves of the equated groups in these schools are compared in this test, the superiority of School DS is seen in the upper half; while in School TS, the reverse is true. It is not clear why this is the case. 20. The differences in arithmetic are negligible between Schools TS and RS. Between Schools DS and RS they are more apparent. The superiority of School DS appears to be almost entirely in the lower half. 21. The differences in the spelling scores are not large, but are consistently in favor of Experimental School TS. No advantage is shown in either the lower or higher halves of the equated groups in this test in the comparisons made between the experimental and control schools. 22. In handwriting there is no evidence of any superiority of the Schools TS and DS over the control schools. CONCLUSIONS Tests reveal conditions, but they do not explain causes. The condition revealed in this study, with practically all the evidence pointing in one direction, is the greater efficiency of the pupils in the experimental types of schools than of those in the control type in the fundamentals of the curriculum as measured by the tests given. Though this difference in efficiency was small, the question arises as to its cause. There is no evidence that it was due to the control type of school having a larger foreign population, a more retarded pupil enrollment, a more irregular attendance, pupils with lower mental ability, a significantly less amount of time in the classroom spent on any of the subjects in which the pupils were tested, oversize classes, less supervision, a body of teachers less experienced or more poorly trained, or, finally, a lower social status of the community from which the pupils came. If it is not due to any of these factors adversely affecting the efficiency of the pupils in the fundamentals of the curriculum, the question occurs as to whether there is something about the 48 Part-Time Types of Elementary Schools organization of the experimental schools which tends to raise the efficiency of the pupils in those schools. Does the very shortness of the time in the classroom create a sense of urgency about the situation which makes for a more direct and effective method of teaching these fundamental skills? May the child acquire much more of these skills in a day than he ordinarily does? Because of the slightly higher standing in this investigation of the pupils in the duplicate type of school organization, the question is prompted as to whether the 225 minutes per week spent in the auditorium-a characteristic feature of this type of school -are not a contributing factor to the higher efficiency of its pupils. In the auditorium are gathered a number of classes at one time. By the less formal atmosphere and the detachment connected with the exercises in the auditorium are the results gained in the classroom augmented in a way that does not occur in the other types of schools? In the mind of the writer the investigation revealed the need of a testing program in the school to give standards and to check up on the progress being made by the pupils. Finally, it should be stated that the findings of the investigation are in favor of the experimental schools, with very slight differences between the two types of these schools. Though in the equivalent-groups method of treating the data the differences between the schools did not reach 4.4 times the P.E., which we set as the standard of practical certainty required for this investigation, yet some of the differences are large and they are supported by the findings of the treatment of the data by the F technique. With the exception of handwriting all the differences are in one direction. These results, as was said, are confined to the fundamentals of the curriculum as shown by the tests given. It must be recognized, however, that there are many other important factors which merit careful consideration in such a study. To do so adequately would require resources not at the command of the writer. But so far as achievement in the fundamentals of the curriculum is concerned, the weight of the evidence is that the experimental types of school organization give the better results in the grades included in this study. BIBLIOGRAPHY AYER, F. "The Curriculum-Organization." Mimeographed Report to the Board of Education, Seattle, 1923. BACHMAN, F. P. Problems in Elementary School Administration. World Book Company, Yonkers, 1915. BUCKINqHAM, B. R. Seventeenth Annual Report of the New York City Superintendent of Schools, 1914-1915. Survey of the Gary and PreVocational Schools in New York City. COLLINGS, E. An Experiment with a Project Curriculum. Macmillan Company, New York, 1923. COURTIS, S. A. The Gary Public Schools-Measurement of Class-room Products. General Education Board, New York, 1919. ENOLEHARDT, N. L., and EVENDEN, E. S. Survey of the Schools of Lawrence Township, Mercer County, New Jersey, 1921-1922. FOOTE, JOHN M. A Comparative Study of Instruction in Consolidated and One-Teacher Schools. Journal of Rural Education, 2:8, 1923. FROST, N. A. Comparative Study of Achievement in Country and Town Schools. Teachers College Contributions to Education, No. 111, New York. IIAGERTY, M. E. Rural School Survey of New York State. A Report to the Rural School Patrons, Ithaca, New York, 1922. MCCALL, W. A. How To Measure in Education. Macmillan Company, New York, 1922. MCCALL, W. A. How To Experiment in Education. Macmillan Company, New York, 1923. NIFENECKER, E. A. Pupils' Progress through the Grades. Board of Education, City of New York, Bureau of Reference, Research, and Statistics, New York, 1922. OTIS, A. S. Statistical Method in Educational Measurement. World Book Company, Yonkers, 1925. PINTNER, R. The Mental Survey. D. Appleton and Company, New York, 1918. SPAIN, C. L. The Platoon School. Macmillan Company, New York, 1924. STEVENSON, P. R. Smaller Classes or Larger, A Study of the Relation of Class-Size to the Efficiency of Teaching. Public School Publishing Company, Bloomington, Ill., 1923. STRAYER, G. D., and THORNDIKE, E. L. Educational Administration. Macmillan Company, New York, 1913. TAYLOR, J. S. Duplicate Schools in the Bronx. Board of Education, New York, 1916. 49 VITA FRANCIS MELVILLE QUANCE, the author of this dissertation, was born at Elimville, Ontario, on January 7, 1883. He received his early education in the public schools of Western Ontario, graduating from the Guelph Collegiate Institute in 1900. He attended the University of Toronto from 1910-1912, and the University of Alberta from 1912-1915, from which university he received the degree of Bachelor of Arts in 1914 and the degree of Master of Arts in 1915. He was a student at the Summer Session of the University of Toronto in 1918 and of Queen's University in 1919; student at Teachers College, Columbia University, Summer School, 1921, during the year 1921-1922, and the Summer School, 1924. -"iSa^.~ ': S....~r:..,........ e-I'N s?^J~,H.I,;.:. |S 'lv 1 "pa.t - ' "'ml * C; IsfW! f' "t:'S r. ~~ V^'Cr; \'^^7 '^-^*1'.- -:.*1;7: ^^ ^ ^ ^^t^^., -,^ ^ ~..:~',.!.,,;... ^.-RB, *..:..,-,: A ^ f. -^..(,r'., ~~-YP-,.-. ^H~ ^'-:H ^-i. ^'~H —.I:t^.^ ^f ^~~.: r-~^ 'Y^ v^^y: w'^Y^ ^^ g Ci;.- $;'* -I,.;P ':~"i %' '`. X r":^^ ^ - w libiwo ^ i '!I I^ "" " s e * H:)(Is^ 0 ) 4 1)<* ~~~~ IA1~~~~~~~~~~~~~ ^ Y ^\^^\^ ^ \^^ ^~/ \!^ \S^ / ~rt *."a~ -U i;.9; I o t^ i~ - o 1 ( S@ so ( \^~/ / ^^ p^ At A, Cecil f4JA\ (two A l:-~-B ~ ~ K 'F~ f. 2~'. ~~ i * tf {t"~~- ^41-,i Qi^ ~ i7~- ^ b -- >^*BP'w"^.23 f-^^ /%^ /^% /% -<* '^'*'S^' Ss^Sfg I sfe'~; — 'l f ao i 'l - ^^^V^~v ^" 3,\ 4~1 Mr, MKi; M ~~~ -~-r~.;;l:;,, ~i~%,I g ~2~-~~~~~~~~~~i ~ ~ i ~ ~ ~ r~~Y &:~~~~~!~~iX. 9 raq Fe, ~. _X` i~::~ i~~i~j~F':~ilP~7c~~f:~':~'A'w ~. s_"::cb~i~~~~4. 0~.i4~~~~~~~~~~~~.~ w: Z-l:Z