arVf" 8761 ^%^ ^.^.i^. ^^^^ i - is * i«^ ■;S^%f -^ V- i^"?^ Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924031784451 I DEPA^RHTMENT OF PVBLIC INSTIWCTION SPELLING EFFICIENCY IN THE ' ■ ; OAKLAND SCHOOLS A REPORT OF THE OAKLAND SPELL- ING INVESTIGATION OF OCTOBER,, 19 14 BY J. B. SEARS. ASSISTANT PROFES- SOR OR EDYfATION. LELAND STAN: FORD JVNIOR UNIVERSITY / BOARD OF EDVCATION BVLLETIN NVMBER 9 NOVEMBER 1915 ii a SPELLING EFFICIENCY OAKLAND SCHOOLS A REPORT OF THE OAKLAND SPELLING INVEST- IGATION OF OCTOBER, 1914 By J. B. SEARS, Assistant Professor of Education Leland Stanford Junior University. Price, Twenty-five Cents PUBLICATION NO. 1. BUREAU OF INFORMATION, STATISTICS, AND EDUCATIONAL RESEARCH. WILFORD E. TALBERT, Director Tune, 1915. INTRODUCTORY NOTE. For the past two years the principals of Oakland have been active members of an organization known as the "1915 Club" which was conducted by Prof. C. E. Rugh of the University of California. The Director of the Bureau of Information, Statistics, and Edu- cational Research, a new Bureau created in July, 1914 for the pur- pose of promoting the scientific -study of Education, took advantage of the opportunity this organization afforded, and led in the forma- tion of a club composed exclusively of school principals. This club is a self-governing body, with membership purely voluntary, and attendance on the part of supervisors or other non-members is only by invitation. The accompanying report represents the results of the first of the club's efforts to realize its purpose. In September, 1914, Professor J. B. Sears of Stanford Univer- sity was invited to address the Club on the subject of "Possible and Profitable Lines of Investigation for School Principals." The result of his inspiring talk was the conducting of an extensive study of the teaching of spelling in the elementary schools throughout the city, with Professor Sears as Director of the investigation. The following pages constitute his findings as reported to the Principals' Study Club. WILFORD E. TALBERT. Oakland, California, June 25th, 1915. Ill TABLE OF CONTENTS. PAGE Introductory Note by the Director of the Bureau of Information, Statistics, and Educational Research Ill List of Tables , V List of Diagrams VI Author's Preface VII CHAPTER I— INTRODUCTION. Purpose and plan of the investigation — How these plans were car- ried out — The nature of the test given — Special problems — Scope of the. test — Scoring the papers — The study of the results 1 CHAPTER II— SPELLING EFFICIENCY BY GRADES, BY SCHOOLS, AND FOR THE CITY AS A WHOLE. The city as a whole — Comparison of schools — Comparison of grades and classes — Individual standings by grades — A study of grade 2 — Summary and conclusions 6 CHAPTER III— SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES. The Problem — Age, grade, and spelling efficiency — Sex and spelling efficiency — General ability in school work as judged by the teacher, and spelling ability — Summary and conclusions 24 CHAPTER IV— SPELLING EFFICIENCY AND. SOCIAL DIFFER- ENCES. The nature of the problem — Father's occupation and spelling effi- ciency — Children's occupational ambitions and spelling efficiency- Father's nationality and spelling efficiency — Influence of home lan- guage on spelling efficiency — Summary and conclusions 41 CHAPTER V— THE ADMINISTRATION AND TEACHING OF SPELLING. The problem — Oakland's policy affecting spelling — Time allotment and spelling efficiency — Time allotment compared with that in use in other cities — Relation of time consumed to achievement in test — Content of the course in spelling — Length of lesson assignment — Methods and devices in use in the teaching of spelling — Con- clusions 61 Appendix A — A Suggested Reading List for Teachers and Principals.... 78 IV LIST OF TABLES. TABLE PAGE 1. Word lists 3 2. Number and grade distribution of children taking the tests 4 3. Standings by schools and by grades 7 4. Distribution of grades receiving highest and lowest averages 10 5. Showing average for all children in the city by grades, and the high- est and lowest standing made in each grade 11 6. Distribution of standings of individual pupils by grades 14 7. Comparison of standings for grades 2A and 3B 19 8. Age-grade distribution of children tested in grades 3 to 8 23 9. Age-grade distribution in per cents 26 10. Age groups and spelling ability by grades -28 11. Showing by grades the standing of three age groups as compared with the averages for the grades from which they were drawn 30 12. Grade averages by sex 33 13. Showing the distribution of standings of 2644 children by occupa- tional or social groups •. t2 14. Occupational distribution of the 7th and 8th grade children and their standings by groups 46 15. Showing the per cent of children of native and of foreign parents who receive the different standings and for those of foreign par- ents separately 51 16. Comparison of errors made by children from English speaking homes with those made by children from foreign homes 55 17. Comparative distribution of spellings among different nationalities — how each word was spelled and by what per cent of each group 57 18. Distribution of time allotments by different classes and grades for study, for recitation, and for the two combined 65 19. Showing the distribution of classes in each grade with respect to the number of spelling periods per week 68 20. Comparative standing of ten cities with respect to the per cent of total school time devoted to spelling 68 SI. Relation between minutes of time per week spent on spelling and the standing made in the test. (39 classes from 16 schools) 69. 22. Showing the chief sources, other than text, from which spelling lists are selected, and the number of teachers by grades who make use of each of these sources 71 23. Showing by grades the variability in length of daily lesson assignment 72 24. Distribution of teachers with respect to methods and devices used in teaching .spelling 75 LIST OF DIAGRAMS. DIAGRAMS PAGE 1. Entire city by schools, showing average for school, highest grade and lowest grade in each school '. 8 3. Distribution of grades and classes for city as a whole 11 3. Distribution of standings of pupils by grades, showing medians 15 4. Distribution of standings by per cents for all children in grades 3 to 8, showing median 17 5. Distribution of class averages: grade 2B, 18 classes, 2A, 10 classes 19 6. Distribution of standings of grade 2 by per cents 20 7. Age-grade distribution in per cents, showing accelerated, normal, and retarded groups 27 8. Spelling ability by grade and by age groups 29 9. Relative standings of boys and girls by grades...- 33 10. Comparative standings of boys and girls in each grade by schools 34 11. Showing the average standing of five brightest and five poorest pupils in each class (based on teachers' judgment) compared with the average for their respective schools 37 12. Showing the distribution of the standings of 2644 children with re- spect to their fathers' occupations 43 13. Distribution of 7th and 8th grade children with respect to their choice of occupations. 3079 in all 45 14. Distribution of fathers of the third and eighth grade children by nationalities 50 15. Showing the per cent, of children of native and of foreign parents who receive the different standings and for those of foreign par- ents separately 52 16. Comparison of foreign groups with their grade averages 53 17. Distribution of study, recitation, and study and recitation time com- bined, by classes 67 18. Showing relation between class standings and the amount of time used for spelling 70 VI PREFACE. The following is a report of the study of spelling efficiency in the schools of Oakland as undertaken in co-operation with the Principals' Study Club of that city. The report has attempted to get at the facts regarding the hand- ling of spelling, and it presents these in as much detail as the prac- tical aims of the study demand. For the most part, the results are highly commendable, but at points a fair interpretation must neces- sarily be critical. The report has sought to make a perfectly frank statement of conclusions, whether favorable or unfavorable, and where it is beHeved that change should be made, definite lines of procedure have been suggested. In accordance with the original plan, the data upon which this report is based was made use of by a small group of my advanced students in connection with a course in elementary education. Much of the data so handled was of use directly or indirectly in the prepa- ration of this report, and I wish here to acknowledge with thanks such assistance from : Miss Leela Kinnear. Mrs. Mary Chamberlain. Miss Myra McHale. Miss Ora Hooker. Miss Mary Ketman. Miss Genevieve Schellbach. Miss Ottilie Hoernig. Mr. George Bettin. Miss Hazel R. Davis. Mr. Will A. Thompson, Jr. and especially from Miss Catherine Cavanagh, who aside from class work assisted in tabulating much of the data for Chapter V. Thanks are also due the Stanford and California University students who assisted in giving the tests, and to the teachers, princi- pals, and other school officers for making available such information as was essential to the study. Finally, I wish to express my hearty appreciation of the constant help of Mr. C, B. Crane, President of the Principals' Study Club, and of Mr. Wilford E. Talbert, Director of the Bureau of Informa- tion, Statistics, and Educational Research, without which the study could not have been made, and also for the excellent spirit of co- operation shown by the members of the Club which has made the preparation of this report a pleasure. J. B. SEARS. Stanford University, California, June 25th, 1915. CHAPTER I. INTRODUCTION. PURPOSES AND PLANS OF THE INVESTIGATION. The following plans for a study of the present state of efficiency of the teaching of spelling in the schools of Oakland were worked out in co-operation with the Principals' Study Qub of that city, and the investigation was .b©g«ti on Oct. 20, 1914. The purpose was to make a complete survey of this branch of the elepientary school curriculum, covering a test of spelling effi- ciencynn grades 2 to 8 inclusive, in all schools, as well as the ad- ministration, supervision, and teaching of the subject. '^ To these ends, 1st, a written spelling test was given to the chil- dren of these grades, for which the following printed form was provided. SPELLING BLANK. Oakland Spelling Test. October , 1914 1. No 2. Name 3. Age (yrs. and months) 4. Grade 5. Language spoken at home by father and mother 6. Father's occupation 7. Father's nationahty _ 8. What occupation do you want to follow when you are grown ? 1 2 3 4 5 9 10 On this form each child was given a number as above. The best pupil in the class, "best all-round pupil, not only in spell- ing but in general school work of all kinds," was numbered 1, the second best, 2, and so on for 5 pupils. The other children were numbered in any order till the last five, who were to include "the five weakest pupils in the class," "weak in general school work, and not spelling alone." The purpose in gathering the information called for here was not only to throw as much light as possible on the problems of spelling, but to have at hand for similar studies in other lines later. 2nd, another printed blank, as follows, was placed in the hands of each teacher of spelling, who filled it out under the *The Beulah school of 20 pupils divided among several grades was not included in this test. 2 OAKLAND SPELLING INVESTIGATION TEACHERS' INFORMATION BLANK. Oakland Spelling Test. OCTOBER , 1914 1. Name Grade No. Tested.... 2. Minutes per week given to f Study of Spelling _ ( Recitation of Spelling - 3. Number of Recitations per week 4. Do you advise or require home study of spelling? 5. Do you teach spelling in connection with any other subjects, as reading or language? 6. Do you take spelling lists from spelling test exclusively?—. 7. Mention any other sources from which you make up lists.. 8. How many words do you assign per lesson? 9. Explain as fully as you can the method or methods by which you teach spelling 10. Describe carefully the devices you make most use of in teaching spelling in this grade 11. Name the best books and articles you know of on the teaching of spelling 12. What subject and grade do you like best to teach?.. directions of the principal. Through this blank it was hoped to get as much information as possible upon the general place of spelling in the curriculum, its connection with other subjects, the time it consumes, and the methods by means of which it is taught. In ad- dition to this, conferences vs^ere- held wit-k the supervisor of Primary grades, with the Principals' Study Club, and with the Director of the Bureau of Information, Statistics, and Educational Research, and access had to such printed materials as bore on the subject. HOW THESE PLANS WERE CARRIED OUT. In order that the tests might be given under uniform conditions, and as nearly as possible under normal schoolroom conditions a time schedule was arranged by Mr. Talbert of the Bureau of In- formation, Statistics, and Educational Research so that all classes of a given grade would take the test at the same hour of the day — lower grades in the morning, and upper grades after noon, — and without knowing that the test was in any way unusual. Some sixty assistants, advanced students of education in Cali- fornia and Stanford universities, who had received both personal and printed instructions as to how the tests were to be conducted, were placed in charge — one in each room during the time of the INTRODUCTION 3 test. In each building, one, a chief assistant, co-operated with the principal in settling matters of detail. In addition to the directions given to assistants, separate uniform instructions were issued to all principals and teachers, and through the careful planning of the principals the regular routine of the school day was broken very little. The words were pronounced clearly and slowly by the regular classroom teacher, with little or no explanation. The children wrote on the blank provided, the information part of which had been previously filled out: items 1, 2, and 4 by the teacher; 3, 5, 6, and 7 by the child with the teacher's help, and 8 by the child alone. These papers were then collected, put into an envelope and labelled by the assistant as to school, grade, room, teacher, time test began, and with a full statement of any irregularities or special conditions influencing the test. All the envelopes from one building were finally placed in the hands of the chief assistant for the building, who delivered them at the office of the Bureau of Information, Sta- tistics, and Educational Research. Here also the principals sent the teachers' blanks which had previously been filled out. THE NATURE OF THE TEST GIVEN. The test given was that devised by Dr. Leonard P. Ayres, which he used in the Springfield Survey*, which was later used in the Butte, Montana, Survey**, and since embodied in a complete spell- ing scale.*** It is composed of seventy words, ten words for each grade, 3 to 8 inclusive, shown in table 1. TABLE L Word Lists. GRADE 2 GRADE 3 GRADE 4 GRADE 5 1 foot 1 fill 1 forty 1 several 2 get 2 point 2 rates 2 leaving 3 for 3 state 3 children 3 publish 4 horse 4 ready 4 prison 4 o'clock 5 cut 5 almost 5 title 5 running 6 well 6 high 6 getting 6 known 7 name 7 event 7 need 7 secure 8 room 8 done 8 throw 8 wait 9 left 9 pass 9 feel 9 manner 10 with Public 10 Tuesday Schools of Springfield, 10 speak Illinois, "an educational 10 flight *The survey' ' Division of Education, Bussell Sage Foundation, E 137, N. T. 1914. **Beport of a Survey on the School System of Butte, Montana, Board of Edu- cation, Butte, Montana. ***A measuring scale for' ability in spelling, Leonard P. Ayres, Eussell Foundation, N. T., 1915. OAKLAND SPELLING INVESTIGATION GRADE 6 GRADE 7 GRADE 8 1 decide 1 district 1 petrified 2 general 2 consideration 2 tariff 3 manner 3 athletic 3 emergency 4 too 4 distinguish 4 corporation 5 automobile 5 evidence 5 convenience 6 victim 6 conference 6 receipt 7 hospital 7 amendment 7 cordially 8 neither 8 liquor 8 discussion 9 toward 9 experience 9 appreciate 10 business 10 receive 10 decision These words were chosen from a large list used by the Division of Education of the Russell Sage Foundation in an investigation of the spelling ability of children in elementary schools. Taking the 1000 words, found by a previous exhaustive investigation to be most commonly used in writing, these words were made into lists, with which children in nearly 100 American cities were tested. On the basis of these tests the words were then classified in groups by grades, each word being placed in the grade wherein it had been correctly spelled on an average by 70 per cent, of the children. That is, the 10 words for grade 2 were spelled correctly by 70 per cent, and incorrectly by 30 per cent, of second grade pupils in these 100 cities, and similarly for each of the other grade groups. Thus the standard established by this exhaustive test of each word comes to be 70 per cent. This not only means that each of the 10 words in grade 3 is equally difficult for children of grade 3 (and similarly of the words in the other groups for their respective grades), but that the words in grade 3 are just as difficult, and no more so, for the 3rd grade children as are the words in the 4th grade for the 4th grade children, the 5th grade words for 5th grade children, etc., for all the grades. This means that the children of Oakland are being measured against the children in these other cities, as well as against each other, and will be classed as high or low accordingly as individuals, classes, grades, schools, and the city as a whole make above or below 70 per cent in the test.. SCOPE OF THE TEST. This test was given in 40 of the 41 schools in the city, to a total of 13,985 children, by grades as follows : TABLE 2. This small number in grade 2 is due to the fact that the 2nd grade test was given in only a few schools and that the test was so evidently a test in penmanship in a very fe\y classes that the INTRODUCTION 5 papers had to be discarded. The records of these 657 are probably- typical of what the better 3nd grade classes can do. They are treated separately everywhere for the reason that it is a deliberate part of the school policy in Oakland to lay little stress on spelling in the first year, not teaching it at all before the last 6 weeks of the year; to lay no great stress on writing in the first year, save for coarse blackboard work ; and to do nothing with written composition before the second half of the second year. SCORING THE PAPERS. The scoring of these papers was done in part by a group of Stanford University students of education, several of whom were teachers of experience, and in part by teachers and principals in Oakland. In the latter case the papers were removed from the envelope, bound by a clip, and given a code number by the director of the Bureau of Information, Statistics, and Educational Research. They were then given out to be scored, after which they were re- turned to their proper envelope. The scores and errors were re- corded on a printed record sheet designed for that purpose, in which the records of boys and girls were kept separate by grades. Where two grades were in one room their papers were placed in two separate envelopes and their standings recorded on separate record sheets. The errors were scored on the record sheet for each word separately, so that if desirable a study of the relative difficulty of the words could be made. Enough of the markings and records were examined with care by the writer to satisfy him that the work had been done accurately. . THE STUDY OF THE RESULTS. After the scoring had been completed, all spelling blanks, record sheets, and teachers' blanks were filed in alphabetical order by schools, and the tabulations and study of the data begun. The first task, (Chapter 2) was the main statistical handling of the children's scores in the tests. These results are presented by schools, by grades, and for the city as a whole, accompanied by diagrams and tables, showing the complete distribution of the chil- dren of each grade by the scores they made in the test. The third chapter deals with the influence of individual differ- ences in age, sex, and general school standing ; the fourth with the influence of such social factors as the father's occupation, nationality, the home language, and the child's occupational ambition. Chapter 5 presents the results of the tests in the light of the administration and teaching of spelling, dealing with the special matters of time dis- tribution, the lesson assignments, teaching methods, etc. CHAPTER II. SPELLING EFFICIENCY BY SCHOOLS, GRADES, AND FOR THE CITY AS A WHOLE. GENERAL STATEMENT OF SCORES. As stated above, this investigation includes a total of 12,98.5 children, of grades 2 to 8 inclusive, and was made in 40 of the 41 schools of the city, on October 20, 1915. Table No. 3 presents in detail the standings of children in grades 3 to 8 inclusive, by schools, and by grades. For the purposes of this study each school is given a number, by which it is known throughout the study. The Key for this numbering is left for use of school officials in the office of the Bureau of Information, Sta- tistics, and Educational Research. Grade 2 was not included in this table for the reason that only a part of the principals considered the test possible for this grade, both from the standpoint of writing, and from the standpoint of spelling. See also Chapter I, section on "scope of test." The results ob- tained, however, are presented underneath the table in their ap- propriate places, and will receive further separate study later. SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. TABLE 3. Standing by Schools and by Grades.* School Grade VIII 1 78.5 79.3 86.3 82.2 82.0 66.1 2 81.1 84.7 83.1 85.0 78.8 67.2 3 72.2 78.4 80.5 89.7 79.9 67.3 4 70.5 77.1 82.2 87.6 77.8 60.6 5' 'si"3 88.5 79.5 81.9 6 78.4 81.7 83.3 84.4 75.4 73.5 7 sT.'s 87.4 80.4 75.2 59.6 8 73.1 73.7 82.1 80.1 77.8 69.7 9 8b"2 71.4 85.9 76.9 67.5 10 76.2 VII 77.4 VI 77.9 V 85.3 IV 74.2 Ill 66.4 School 1 80.8 80.2 77.1 75.2 82.8 79.8 76.5 76.5 73.0 76.8 Grade II 1 72.0 77.2 80.0 57.7 51.2 78.2 School Grade VIII.. VII.. VI-. v.. IV.. Hi- ll 75'.7 81.3 78.6 88.8 77.7 75.3 12 73.9 81.2 79.1 75.3 66.2 58.1 13 81.9 81.4 80.0 88.2 78.4 75.9 14 68.9 81.4 79.0 86.3 71.4 58.7 15 80.8 78.6 79.4 84.8 76.0 60.7 16 78.7 81.7 84.8 84.7 74.7 59.8 17 71.9 74.9 72.0 80.2 77.3 70.8 76.2 80.7 19 74.9 84.6 72.0 81.9 71.8 67.4 20 80.7 84.4 75.8 81.5 76.0 63.8 School ■■- I 76.7 I 71.5 I 80.4 | 74.7 | 74.2 | 78.9 | 74.4 | 79:0 | 71.1 | 76.8 Grade II- .L 87.5 I _L _L _L JL Grade II I 73.3 65.4 I 74.7 School 21 22 23 24 25 26 27 28 29 30 Grade VIII 70.0. .75.1 83.9 79.8 72.0 65.6 VII ,_,. 76.7 «2.5 71.0 75.0 77.8 75.8 ^ VI 77.7 80.0 68.0 77.3 71.8 69.8 71.7 89.7 84.2 82.7 V 80.2 93.5 73.9 79.5 68.9 75.3 86.3 88.1 84.4 IV 81.2 75.2 70.3 76.3 72.4 66.8 78.2 80.0 75.5 66.7 Ill 74.3 74.8 60.2 69.7 61.5 59.1 57.7 74.8 68.0 74.6 School I 78.3 I 79.7 I 70.8 I 76.8 I 68.3 | 70.8 I 74.2 | 78.5 ] 78.4 I 74.1 67.7 I 52.5 I I 52.0 | 64.0 School 31 32 33 34 35 36 37 38 39 40 Grade VIII.— 68.6 VII _ 72.8 79.4 VI 90.5 81.3 71.7 V 82.9 79.8 87.6 80.0 78!6 87.4 76.8 88.7 72.8 70.0 84.6 63.3 80.8 74.9 75.5 70.7 IV-_ in 68.1 75.1 65.0 49.2 63.5 62.1 37.5 63.9 70.2 School - I 74.5 I 79.3 I 70.0 I 71.5 I 77.6 | 68.3 I 74.1 I 63.7 I 71.1 I 70.2 Grade II- .| 37.9 I _L _L _L _L _L _L I 70.8 I *Average for city^77.40 Average for city=76.53 (grades 3-8). (grades 2-8). It will be seen from this table that the standings by schools vary from 82.8 in school No. 5, to 63.7 in school No. 38, but that only three schools fall below the standard of 70 per cent, correct. Sim- ilarly the schools may be compared by grades, and likewise the grades within a single given school. These and other features of this table appear more clearly in diagram No. 1, which shows at a glance the general distribution of standings by grades in all schools. The upper line in this diagram OAKtAND SPELLING INVESTIGATION O K u CO ( o Ski Si s ^ 1 J S3 1 h In ,, ra ^ _ n •^ vS^ ^^^^i^^^^^^^^ « VH^ XS] n ... ^ ^ i M J^^^ ^f o R o F?w^ ^\NH n 8 8 10 ^ n 1^^. n 4* ^ ^— , ^ « ^ < ^^i^ (0 p H 4) .s x>- 5 S 3 a ■* to J) t f n ^s^s r n S V 9 3 S vS S>v^ « in \N^ » ■$51 « 5 n ^^>:^ w KSv to 2 m ^^&$; ^ 10 n o y^ » in ^^ 1 n ^?^ 10 ■^ .t^ in m o ^ ^ 1". n ") ^^ O n c^ V to i , 1 i9 S 10 ^r ST 8^ ■ 'fe SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 9 represents the average standing for the best grade- in each school, and the number of that grade is indicated for each school by a figure immediately below the number of the school. In like manner the grade receiving the lowest standing is indicated for each school by the lower line of the diagram, below which is the row of figures indicating which grade has received such standing. Between the upper and lower lines is a third line which represents the average standing of the schools, while the heavy straight line marks the standing achieved in other cities. On the right, a short horizontal bar marks the line on which the city average would fall if shown on the diagram. Thus each school, as well as the best and poorest grades of the school, may be conveniently compared with achieve- ments in other cities, as well as with other schools in Oakland. Similarly all comparisons may be made with the average for the city as a whole, which for grades 3 to 8 is 77.4 per cent., or with the 2nd grade included, 76.5 per cent. THE CITY AS A WHOLE. At a glance, therefore, it appears that the average for all schools in the city stands 7.4 per cent, higher than the average standings made in other cities, that the line "representing the highest grade averages is uniformly high above the 70 per cent, mark, that the line representing the lowest grade averages is also relatively high, while in only a few cases are either of these extremes strikingly high or low. If the 28 2nd grade classes which took the test were included, this difference would not be materially changed, as will appear later. In general, therefore, the spelling in Oakland is de- cidedly superior. The extent to which this remains true in particu- lar will appear later. COMPARISON OF SCHOOLS. A study of this diagram by schools, at once reveals the zigzag nature of the middle line representing school averages. The fact that the school averages for' 23 schools fall below 77.4, the average for all children in the city, that 3 schools make just the city average, leaving 14 schools to offset the low averages in 23 schools, indicates that the larger schools did the best spelling. That is. school No. 1 with over 700 children tested, makes an average of almost 81 per cent, while school No. 23, with the same number of grades but with only 250 children tested, averages below 71 per cent. The widest difiference between school averages being shown by school No. 5, with an average of 82.8 per cent, and school No. 38, with an average of 63.7 per cent (neither of these schools contain grades 7 and 8). Further evidence that the large schools are responsible for the high city average is seen in the fact that every grade average in threb 10 OAKLAND SPELLING INVESTIGATION entire schools, no one of which includes over 350 children in the test, fall entirely below the city average. By a study of the upper and lower edges of the shaded portion of the diagram one sees precisely this same thing, only somewhat exaggerated. The highest grade averages made by schools 25, 36, and 40, all small schools, are entirely below the lowest grade aver- age made in 7 of the other schools, 5 of which are among the largest schools in the city. Not only is there a marked difference between schools in these respects, but also in respect to the wide differences in the range occupied by the grade standings in different schools. This is quickly seen by comparing the width of the shaded portion of the diagram at different points. The grades in school No. 18"make average standings ranging from 66 per cent, in grade 3 to 86 per cent, in grade 6 ; school No. 38, from 37 per cent, in grade 3 to almost 81 per cent, in grade 6 ; while the range in school No. 5 is only from 79 to 81 per cent. ; and in No. 25 from 61 to 72 per cent. The ex- treme range in average grade standings for all the schools is from a difference of 2 per cent, in school No. 5 to a difference of 44 per cent in school No. 38. A glance at the width of the shading for different schools impresses one with this rather wide difference in the efficiency of different grades in the same school. This will be further examined later. Meantime attention is called to the two rows of figures indicating the grades receiving the highest averages in each of the 40 schools. These are brought together in Table 4, which shows that the poor spelling, so far as grade averages TABLE 4. Grade 3 4 5 6 7 8 No. of schools in which it receives highest average.... 3 21 8 5 2 No. of schools in which it receives lowest average 33 1 4 count, is found mainly in the 3rd grade, while the 5th grade most often ranks highest. In no school does the 3rd grade rank highest, nor do grades 5, 6, or 7 receive the lowest average in any school. With all the above differences between schools there is this similar- ity, viz. that the 3rd grade uniformly ranks low, and the 5th grade high. COMPARISON OF GRADES AND CLASSES. This same fact is emphasized further by Table 5, the facts for which are made clearer by diagram 2. Here the standings for all the 8th grade children throughout the city have been averaged, and SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 11 TABLE 5. Showing Average for All Children in the City by Grades, and the Highest and Lowest Standing Made in Each Grade. GRADE VIII GRADE VII GRADE 1 VI 1 GRADE V GRADE IV GRADE III GRADE II Highest Class Ave 86.3 76.3 64.3 88.7 79.7 71.0 90.7 80.2 60.0 93.6 84.7 66.1 87.3 75.9 • 58.7 88.8 66.7 23.3 87 5 Average for grade Lowest Class Ave. . 60.4 23 5 similarly for the children of each of the other grades, figures for which are shown in the second line in Table 5, and graphically by the central line in diagram 3. From grade 8 this curve rises grad- ually to grade 5, and then descends more abruptly to grade 2, which is included in this diagram for the sake of comparison. From this DIAGRAM 2. Distribution by Grades and Classes for City as a Whole. VTTT snr 'HI loo; HIGHEST CLASS AVERAU AVERAG.E eon e.RJVOE 70 LOWEST CLASS AVEKACe nr HE H GRADE 12 OAKLAND SPELLING INVESTIGATION it is seen that averages for grades 5, 6, and 7, stand higher than the average for the city as a whole (grades 3-8), and that averages for grades 2, 3, 4, and 8 fall below the city average, the widest dif- ference between grades being 18 per cent, which is the differenc between the averages for grades 5 and 3, or 24.3 per cent, if grade 2 be counted. Eighteen per cent, is a wide difference, but when we pass above grade 3 this difference falls immediately to less than 9 per cent, which fact further places the responsibility for the poor spelling on the low grades. Measuring the grades against each other, and against the city average, we see that while the child is doing poor work in grades 2 and 3 (poor also as measured by achievements in other cities), his rate of improvement is rapid till the fifth grade is reached, but that this rate is not only not main- tained, but that it materially decreases through the subsequent three years. An explanation of this attainment for the 5th grade may lie partly in the fact that a careful review of spelling is made during this year. A study of averages, however, is only a rough method at best, and may not reveal the most important differences in the efficiency of the different grades. In this same table is presented the records of the best and of the poorest class averages found throughout the city, for each grade. In the diagram these facts are shown by the upper and lower margins of the shaded surface. This brings out a few characteristics of the work in the different grades which does not appear in a -study of gfrade averages alone. The variability of classes within each grade is quite an important item, and may be very effective in helping to show up some of the important admin- istrative and teaching problems. The less variability among the classes of a given grade the greater the probability that a consistent policy is being pursued in the teaching of spelling in that grade; while great variability would tend to indicate a lack of uniformity in the administration and teaching. This does not mean to imply that absolute uniformity in results is either possible or desirable. A question is fairly raised as to the desirability of having a uniform plan for handling this particular subject, not, of course, for all grades, but for all classes of a given grade, regardless of the nature of the community or school conditions. Clearly various factors would enter to make such uniformity unwise and impracticable, and similarly uniform results would be difficult to obtain. However, teaching standards should not be confused with the nature of the content taught. Differences in content for different types of com- munities, which at most would be very slight, need not demand wide differences in the standard maintained. No theoretical answer is offered for this question here, but at- tention is directed to the facts presented in our table and diagram, which show that this fact of variability of classes within a given grade is widely different in different grades. The fifth grade does SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 13 not occupy first place now, though it still stands above the earlier grades. The classes in grade 7 are least variable and those in grade 3 most variable, even including grade 2. The class standings be- come more and more uniform until grade 7 is reached, from which grade 8 declines slightly. The range in class standings is so great here that it is clearly evident (though not proved by this test) that the class represented by the lower margin of the shading for grade 3 is as much below the class represented by the upper margin of the same grade as the 3rd grade is below the 4th grade. While this overlapping between grades cannot be measured quantitatively here, it is clearly very marked. Just what variability of classes within a grade should normally be expected would naturally depend, 1st. upon whether or not the subject matter of this branch of the curriculum is really organized by years (the subject matter for each year representing a definite amount of work to be done), and 2nd. upon the extent to which promotion in spelling is made only upon the basis of having attained a certain standard of efficiency, which is uniform for all grades and schools, in the subject matter outlined for the year. If there is laxness in maintaining a standard for spelling promotions, and a looseness in the organization of the subject matter, as between schools, and grades, then a wide variability in class averages would naturally result. If promotions must be made by years rather than by subjects then naturally some subjects will suffer with almost every child. No system of grading can be made so elastic as to completely do away with the problem of individual differences with- in a class, unless it is based on individual attainment by subjects. Oakland approaches this ideal, in that promotions may be made at any time, and are regularly made semi-annually. Yet, whatever this may have done to affect a better classification in other subjects, it has accomplished little or nothing for spelling." The facts shown by this diagram tend to argue, particularly in ^Note: It is fair to raise the question as to the liability to abuse which an ideally elastic system of grading may possess. When a child is doing well in his class he may easily go unnoticed by his teacher. If he is doing excellent work he is attractive and the teacher may be reluctant to recommend his promotion. If he is stupid, over age, and not so attractive, his promotion is likely to be recommended as early as possible. Thus, without a regular promotion time, annual, semi-annual, or quarterly, the child is less likely, to be pushed out and made to work up to his limit, and the dull and unattractive child is more likely to be pushed out too soon. To shorten the promotion periods from three months to six or three weeks, is to increase the probability that a teacher will decide to keep her bright children one more period — since it will make her room attractive and at most delay the' children but little — and to risk pushing the unattractive ones faster than they should go. This is a feature of short promotion schemes which needs study. The retardation figures presented later tempt one to think that the short promotion periods which have been possible in Oakland may have been a contributing cause to the high percentage of retardation. The statistics on this subject ought to be dealt with by the Bureau of Information, Statistics, and Educational Research. 14 OAKLAND SPELLING INVESTIGATION the early grades, that there is no definite grade standard of spelling efficiency maintained in the city. INDIVIDUAL STANDING BY GRADES. Thus far we have dealt with averages or central tendencies and with coarse measures of the variability of groups. These methods are adequate for certain purposes, but tell only roughly of the dis- tribution of the individual standings of the children. For this pur- pose the actual distribution of scores is presented in Table 6. This table includes over 12,000 individual records, and shows their dis- tribution in detail. The total line is reduced to per cents, for con- venience of comparison. TABLE 6. Distribution of Standings of Individual Pupils by Grades. Grade Standings 100 90 80 TO 60 50 40 30 20 10 No. of Chil- dren VIII VII VI V IV III.. Total 263 384 533 726 246 476 503 554 368| 541 314| 442 242 359 420 358 500 384 17D 242 290 248 407 337 131 148 162 119 234 246 108 93 106 81 144 227 53 55 79 44 94 189 17 15 13 261 8 48 29 I Same by per cents.. |2588|2765|2263 -I 1 I I I .1 21.01 22.5| 18.4 1703 1 1040 759 \ 1 142! 100 1 514 331 182 105 I 13.8 1 8.4 6.2 4.2 2.7 1.5 1 1 3 46 53 1293 1805 2154 2171 2385 2495 12303 100 The first point of interest is seen in the fact that slightly over one child out of every five, made a standing of 100 per cent. That is, the test was too easy, and therefore did not test 31 per cent, of the children in grades 3 to 8. Another 28.5 per cent, of the chil- dren made 90 per cent, on the test, while on the other hand, 15.9 per cent, of the children received a standing of 50 per cent, or lower. Complete distribution of individual standings for grades 3 to 8 is shown for each grade separately in diagram 3. The distribution of the spelling abilities of a group of 2000 children selected at random regardless of grade would be practically SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 15 DIAGRAM 3. Distribution of Standing of Pupils by Grades, Showing Medians GRADE ■mi rZ93 CHIIPREN sRADE^nr ISOS CHILDREN GRADE 5: 2ITI CMIIORCN ioo 90 VO .70 60 00 40 so 90 I GRADE IS. CHILDREN 40 JO to JO O GRADE HE 2495 CHILDREN represented by a probability curve. That is, a very few would re- ceive 100 per cent, in their test and a few zero, the great majority falling around 40 to 60 per cent. A school grade as here used is a selected group, and the words of the test are selected with respect to a definite degree of spelling efficiency. The children of a given 16 OAKLAND SPELLING INVESTIGATION grade are presumed to be about equal in spelling- efficiency, that be- ing the reason for their being in the same grade, and the words for each grade are, by test, equally difficult for many hundreds of chil- dren of the grade in question. We should therefore not expect a chance distribution of the standings for any grade here, containing from 1200 to 2400 children. We should expect a very definitely marked central tendency. By definite arrangement of the test (see Chap. 1) it is planned to make the central tendency around 70 per cent, correctly spelled. An examination of Diagram 3 shows that in but one grade is 70 per cent, the central tendency, as judged by the median, which is indicated for each grade by a thin vertical line marked M. In every case the curve is badly skewed toward the high standings. So the distribution' of individual standings is practically like the the distribution of classes and grades. Even though the curve is skewed toward the high end, yet there is a good- ly showing at the low end as well. To what extent is this marked diversity to be expected under a rea.sonably careful handling of spelling in the schools? Grading, ideally, means grouping children of like efficiency in a given subject, or of like capacity for handling the subject. It is rarely practicable to attain this ideal in even rare cases. First, because grade, here, as in most cities, means two definite groups one-half year apart in efficiency. This would naturally tend to give us a distribution of their standings, which, if plotted on the curve would show two high frequency points in place of one. This nowhere occurs in Diagram 2. Again, in school practice, almost every school promotes a few children for reasons other than those assigned above. There are always the few over-age misfits. Now if there are no ungraded classes to which these can conveniently be assigned, they are pushed forward, and would tend to push our curve over toward the low end. Then again, there is the child who comes into the grade from some other city, or from the country school. He is also frequently very hard to fit into any grade, and may easily appear either far above or far below our central tendency. Then there is the child behind his grade because of illness, and the one behind because he has jumped a grade. All these cases and others like them would tend to flatten out our curve. Yet they do not help to explain the skewed condition we find, say in grade 5, nor, even with a liberal allowance, do they seem to explain why nearly 16 per cent, of all the children in the city receive a standing of only 50 per cent, or lower. The cases included in Diagram 2 have been brought together, and are expressed in per cents, of the total number of scores, rather than in absolute figures, in Diagtam 4. From this we see that a larger percentage of all the children in the city received a standing of 90 per cent, than any other one standing. SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 17 DIAGRAM 4. Distribution of Standings, by Per Cents — for Children in Grades 3-8 — Showing Median. S^d^nAi If the test had been difficult enough to have tested to their Hniit the 31 per cent who made a perfect showing, it looks very much as if we should have gotten somewhere near a chance distribution of the standings of these twelve thousand children. But the test was not difficult enough so we can only speculate, and say that on the average, as compared with other cities, the Oakland work in spelling is decidedly superior. Averages are not full statements of all the findings, however. For this the complete distribution of standings is needed, and from this it appears that the distribution of standings for most classes and grades is not far from what it would be if the children were not graded at all. That is, two children selected at random from the same class or grade, appear to be no more nearly equal in spell- ing efficiency than two children selected at random from two ad- jacent grades. Space cannot be taken to prove this further by citing typical records of classes, but a study of these individual records by classes is quite as confirmatory of this statement as are the above facts. Large overlapping is apparent in every grade, but more so in the early than in the later grades. Through this overlapping the evidence appears to prove that what is true in general with respect to spelling efficiency in the city, is not true in particular. For, after allowing liberally for all condi- tions which must naturally produce a variation from the central 18 OAKLAND SPELLING INVESTIGATION tendency, there is yet too wide a diversity of abilities within a given class or grade. So wide, in fact, that the word "grade" has prac- tically no meaning. This condition raises important problems in teaching and administration which should command attention. A STUDY OF GRADE II. The data for grade 2 are not included as part of the study of the whole city, because tests were given in less than half of the schools, and where given did not always include all the 2nd grade children in the school. Consequently the results cannot fairly be compared with results from other grades. In view of the fact that a definite policy with respect to the place of spelling in the first two years' work is in operation under direc- tion of a special supervisor for the primary grades ; and in view of the fact that the 3rd grade ranks extremely low as compared with other grades, it seems worth while not only to study the data ob- tained from grade 2 separately, but to study it as fully as the data will permit. As to whether it is, or is not, worth while for any 2nd grade, in any city, to be able to pass this test in spelling, is not the question here, the purpose being, first, to present the results of the test, and to analyze them so as to be able to place the responsibility for the results obtained. We wish to say, in terms of the test, how well these 2nd grade classes spell, and not to raise here the question of whether it is desirable to have them spell better. There are 17 schools represented, varying in the number of chil- dren per school from 9 to 110, including in all, 657 children. There are 28 classes represented, 18 being grade .2B (the upper half of grade 2) and 10 being grade 2A classes. In some schools all 3nd grade children are included, in others, only those of the upper half, or 2B. The groups are treated separately therefore and without reference to the school concerned. It should be stated here (see Chap. V) that those of grade 2A had had six weeks' work in spelling last year in grade 1, and about eighteen weeks' work this year before the test was made. ' The upper half had had one-half year more training than this. The class averages for the 2 groups are shown separately in the upper and lower solid lines in diagram 5. The range in these aver^ ages is large, as appears in Table 7. The average for the 2 are shown separately by dotted lines on the diagram, while the average for the two combined is shown by the solid horizontal line. It will be seen that the average for the upper group is 22 per cent, higher than the average i&r the lower group, neither attaining the 70 per cent, made in other cities, and that their combined aver- age falls almost 10 per cent, below it. The zigzag dotted line rep- SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 19 DIAGRAM 5. Distribution of Class Averages: Grade 2B, 18 Glasses; 2A, 10 Classes. • ' I !■ N> »■■ • 'I '\ ft , ^ ,5U- 6y.^^fQT:-S>ifde. 4/1 .a./? resents the standings of the B classes in the schools from which the ten A classes are drawn, each B class record being plotted opposite the A class of the same school. The lines are by no means parallel. That is, the difference between the spelling efificiency in the A and B groups in one school is no evidence of what it will be in another TABLE 7. Comparison of Standings for Grades 2A and 2B. —2nd Grade— — Grade — — Qass Standing — 2B 2A Both Highest 87.5 52.5 69.0 67.5 23.5 47.2 87.5 Lowest 23.5 Average 60.2 school. Perhaps this should not be drawn on too heavily as evi- dence that the policy for spelling is not carried out in the same way in all schools, but it certainly points that way. The distribution of individual scores for the grade is shown by per cents, in diagram 6. It appears that a larger percentage of the children received a score of 90 than that receiving any otlier one ■score. The median here falls upon 70. A comparison of this with diagram 4, which shows grades 3 to 8 in this same way, makes 20 OAKLAND SPELLING INVESTIGATION DIAGRAM 6. Distribution of Standings for Grade 2 by Per Cents. 100 90 9o 70 bo 50 f 030X0 10 O diagram 6 appear very flat. And even when compared with the distribution of standings for grade 3, it appears extremely low, in spite of the fact that the averages of the two are not so wide apart. It is evident then, that grade 2B ranks almost as high as grade 3. \A'hile the grade average is sUghtly lower, its class averages are not so variable. On the other hand, grade 2A, not one class excepted, falls far below by every measure. If the fact that some classes were not included means that they would rank still lower than those tested, then the children in Oak- land learn very little spelling during the first year and a quarter, so far as can be measured by this test. It remains to be added that when the results of grade 2 are added to those for the other grades the averjige for the city is 76.5 per cent., which is .9 per cent, lower by the addition of this grade. SUMMARY AND CONCLUSIONS. The investigation includes 12.985 children from grades 2 to 8 inclusive, in 40 schools, and shows the following results : 1. For the city as a whole, and for grades above the 3rd, the standing is decidedly high, being T.-l^- above the averages in other cities, or 6.5% higher if the 2nd grade is included. 2. The differences between the averages for different schools is pronounced, the high averages being maintained on the whole by the larger schools, and the low averages by the small schools. 3. Similarly there is a wide range in the variability of grade SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 21 averages within various schools, all grades in one school receiving close to the same average, while in another school they range from 37.5% in grade 3, to 80.8%^ in grade 5, grade 3 being uniformly low, and grade 5 unifonnly high. 4. When the standings of all the children in the city are brought together by grades, this same fact is emphasized by the wide range in grade averages. The difference between the averages for grade 3 (the lowest grade average) and grade 5 being 18%, which rises to 24.3% if grade 2 be included. 5. The averages for the classes within any grade represent a still greater variability than we find between grade averages. This range is greatest in grade 3, and least in grade 7. 6. A distribution of individual scores by grades shows the same extreme variability, and this obtains also in a marked way between the individuals in the same class. 7. Over one-fifth of the children were not fully tested because they spelled all the words correctly. 8. About 1 child in 10 spelled less than half the words correctly. 9. The upper half of grade 2 ranks little lower than grade 3, but the lower half of grade 2 shows very low standing. 10. Practically all these showings point to a serious over- lapping between grades, so marked in grades 2 and 3 as practically to destroy the meaning of the term grade. Which means that the high general showing for the city as a whole, and for all grades above the 3rd, does not obtain in particular when applied to classes and individuals. 11. It must be said then, that, as between separate schools, and as between separate grades in the same schools, there is evidenc that no definite standards for administering this branch of the cur- riculum exist. That is-, a child graduating from school No. .5 is probably 12% more efficient in spelling than is a child graduating from school No. 23, and still greater differences exist between other schools. Similarly it would appear that a lower standard for spell- ing efficiency is used as a basis for promoting a 3rd grade child, in most every school, than that used for promoting a 5th grade child. This is undesirable in school practice. It may be that less emphasis, relative to other studies, should be placed upon spelling in one year than in another, but if so, the amount of work to be done should be adjusted accordingly, so that the quality of work will not suffer. 12. If the subject-matter is to be determined, not by a text alone, but also by current use in written exercises, then a systematic gathering and classifying of words from the latter source should be made an immediate object of study for teachers and supervisors. 22 OAKLAND SPELLING INVESTIGATION One or two years of this work throughout all schools would furnish practically all the material that- is to be used from this source. The words so gathered could be classified by grades in the order of their probable need and difficulty and made into supplementary lists to accompany the text. This would help to systematize the subject- matter and furnish a basis for standardizing the work. 13. Efficient work cannot be carried on in a class whose pupils are sixty to ninety per cent, apart in abilities. Such extreme dif- ferences are not rare, and represent a teaching problem which, in most other subjects would be regarded as practically impossible to handle. In spelling, it is, from the standpoint of real teaching, quite as impossible to handle, and is likely to induce many teachers to resort to extremely unpedagogical methods. In such a class a proper assignment for the children of 80 to 100 per cent, ability is an imposition on children of zero and 10 per cent, ability. If pupils can be classified in arithmetic they can also be classified in spelling. 14. That the showing for grades 3A and 3A is extremely poor as compared with our 70 per cent, standard, cannot be doubted. The present policy with respect to spelling, as laid down by the supervisor of primary work, may account for this (for fuller dis- cussion of this point see Chap. V). The 3nd grade in Butte aver- aged 86.2 per cent, correct, the 3nd grade in Springfield made even 70 per cent., while for Oakland its average is 60.2 per cent. That for the upper half of the grade, however, is 69, or practically up to the standard. Somewhat the same condition exists in grade 3, though with some less emphasis. Which indicates that by the mid- dle of the 3rd year the Oakland children average well up to the standard, and then rise rapidly above it. It does not appear, there- fore, that this low showing is of any serious moment, since children below the 4th grade rarely need spelling. Two questions arise, however: what gain is there in deferring the serious study of spell- ing for the first two or three years ; and, what is the loss in having the added burden later? The gain must be in relieving the early years of much rather formal study, and so adding to the content and expression sides of their training; the loss must be in adding to the later grades more of the formal, and so consuming time that should be used otherwise.* The writer believes that the gain is greater than the loss, pro- viding: first, that this opportunity for the early grades exists; sec- ond, that evidence can be given that achievement in reading, lan- guage, etc., in these grades is superior; and third, that the fourth ^It is not proved yet that from the standpoint of spelling alone, it is an advan- tage to delay the teaching of spelling. It may he so much easier for a 3rd grade child to learn how to spell the word horse than it is for a first or second grade child to learn it that it pays (from the standpoint of spelling) to delay. On the other 'hand, we have no proof of the opposite position, hence it is at present impossible to furnish experimental evidence that the practice in other cities is wrong. SPELLING EFFICIENCY BY SCHOOLS, GRADES, ETC. 23 grade work in arithmetic, language, etc., is suffering no handicap because of the extra work in catching up in spelhng. In the light of the achievements "in 2nd and 3rd grade spelling in other cities — which seems to argue against slighting spelling in the early grades — some investigation of reading and language effi- ciency in these grades would be desirable, and is recommended. These need not be xhaustive, but should cover the early grades in typical schools. Further light is thrown on this situation in chapter V, which tends to show that from the point of view of time these grades are not neglecting this subject any more than are the other grades, which argues more strongly still for a study of results in these other subjects. Present practice in any of the school subjects is not sufficiently professionalized, and free from the force of tradition, to make it a final argument for or against a given practice. The comparative method of study has its merits, however, and its showing cannot wisely be ignored till experimental evidence displaces it. 15. The excellent showing which the city has made as a whole must neither blind us to these very genuine problems ; nor must we fail to examine into the reasons for this high showing, as well as into the fact that no standards with respect to the subject-matter and administration of the subject of spelling seem to exist; and as far as possible into the time cost at which, and the method by which, the showing is made. These are problems for succeeding chapters. CHAPTER III. SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES. THE PROBLEM. The above results have been stated in terms of schools, grades, classes, and individuals. Differences between schools, and between grades were noted in interpreting results, but no attempt was made to interpret the results for children in terms of individuality. Each child was counted a child only. Yet we are conscious of the fact that children are not alike, and that their differences are the begin- nings of many important school problems. To say that the city as a whole, or that a class as a whole, made a given standing tells very little. Furthermore, to show the distri- bution of standings for a class, or for a school, does not indicate the causes contributing to that distribution, and hence offers no directions for changing it if it is unsatisfactory. These results are the first step, however, in a diagnosis. But to say that the city made a standing of 77.4 per cent., does not tell us the kind of chil- dren the city had to have in order to make that, nor does it tell us the plan used in making it, nor the amount of time consumed. To make our results of value, therefore, we need to study as fully as possible into the conditions under which such results have been obtained, for an}- investigation, results of which are not to be made use of in school practice is not only professionally useless, but vicious. If further study proves these results to be superior, then we should pursue our present policy with respect to spelling, with the consciousness that we have only to refine this policy and develop skill in its administration. If, on the other hand, our results prove to be inferior, we need to study carefully every aspect of the prob- lem in order to determine what readjustments are necessary. To this end the results have been studied in the light of differ- ences in age, sex, and general ability as judged by the teacher, and the findings are briefly presented below. AGE, GRADE, AND SPELLING EFFICIENCY. The school practitioner need not concern himself so much with the psychological significance of maturity as with the practical effect of maturity upon the problems and the output of the school. When a class or grade standing is presented as the average for children" var\'ing from a few months to four or five years in their ages, it is important that we should be able first of all to state what effect this wide difference in age has had. This is our problem here. Every grade group in Oakland is made up of children of widely varying ages, and this is only a little less true of half grade groups. SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES '25 CO a, o u O 1i n o bD C -F-l •M o s > o 00 2 o ^ CO o £ H 13 O ■t-i 9) a n o i ■ «5 ^J ^ ^ 5'i;s c^i '^^ ^ 5)^1 f< ^452? OJ N, ^ CO^'OO c^ -- ■>» '6^^5 - ■s. V.' c^ ^ so ^^5^" ^ 'o $ >.^ 1^5: '^^ ■- — •^ to Wi, ^ 5 **». W ,^ -s Q ■^ ^ ^^iS - - "^ h (^ "1 5S ^ iQ4^^ Orj ^ -s t-. sa \ en t S f^^. CO Oo 5 *- CO 5? ^ ■^ 5 ^.S bo ^ ^ s ^ to ^^^^^ l^ :t- w o en, ?: 1^ 05 CO S 5^!^ <> ^ c? 3^ ?: ^ 3- § 5*i 5 j3o § Co § ^ to VO Q ^ 5i 1^ ^ 5*3^ S 5 s 2 ^ W ^ 5^1 53- oo CO CM CM 2; 13- § "^ 35^5 s 00 2 In w is ^^ CM CM §: ifi «lo § s:* ^ Q^iS^oS i ?: ■v. :*- ;Si<» '-0 ■^ ^i- PO Ji!^ ^, Jg^ 00 CQ < ■CO CD CQ 00 26 OAKLAND SPELLING INVESTIGATION TABLE 9. Age-Grade Distribution in Per Cents. &RADE ,4CCEi.e/?ATED Aftf/fMAL ff E T A R D ELT> \ Ol/er / yr; /V1- '1^ a j^ra. 3 1.5 ^•Y V.6 4/\ ./ s-.i 5¥.? 31 .lo 16-.% 7.2 S-.H- 4B .3 7.5- ^if.S- 30.LI- /7.5- r.f- Ic.O 5A . / ^.1 3.^,1 sr.q /9.7 ^.3 5-.% 5B .7 7.7 ^S.<^ 30.H- /^.H- /O.O 5-.<^ 6A .0 , 6.^ 3,H-.0 .29.^ 2.1. 'I I2.H- s-.^ 6B .;i 5".^ ^l.if 3:2.2 2S.^ II .7 3.3, 7A .6 ^.7 ZH-.Z 31 .& J25-.7 /O.H ^.7 7B .7 6. a Zt+.l 31. % 2s:i q.s 2,.^ 8A .6 ¥.3 2.2.0 37,3 2.2.5- /C^f 2..>i 8B .¥ 7./ zz.o 32.'S 21.1 S'.h a..o ALL (7.3 6.2 28.6 31.5 20.0 9,1 4.3 . For convenience the age grade distribution of the children included in the test — 3nd grade excepted — is presented in table 8, reduced to per cents, in table 9 and shown graphically in diagram 7. It will be seen that the amount of retardation is high, ranging from above 54 per cent, in grade 3A, to above 70 per cent, in grade 8B, and amounting to 64.9 per cent, for the six grades studied. This means that retardation for the elementary schools in the city is at least 60 per cent., if first and second grades were counted. The amount of retardation for each child is shown in table 8, and by per cents, in table 9, from which it appears that from 28 to 37 per cent, of the children are retarded one year, that from 12 to 27 per cent, are retarded two years, that from 5 to 12 per cent, are retarded three years, and that from 2 to 6 per cent, are retarded four years or more. An examination of the accelerates shows that from 4 to near 8 SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES 27 < en per cent, are one year ahead of their grades than one year ahead. It is interesting to retardation increases gradually from the grade. A comparison of the average standings grades will show roughly the influence of tained. These facts are given in detail graphically in diagram 8. The left hand represents the average standing for the , while very few are more note that the amount of low 3rd to the high 8th of these three groups by age upon the results ob- in table 10, and shown column in this diagram accelerates, the middle 28 OAKLAND SPELLING INVESTIGATION TABLE 10. Age Groups and Spelling Ability by Grades. GRhDE STA/VD/A/as BY G ROU PS Acceleva-fces A/or-m^L^ Se.t'avt/s 8 B §7.7 ?^.0 7S.3' 8 A S^^.O T^ .7 ^^.f 7 B ^^.ii g'7.3 ^.7 S B ^<5-,7 ^ f .^ ?2,.'3 5- A "^S.H- ^7.0 7^.7 4 B ?^.? S'O.S 7(0.7- 4- /\ 7 9. 1 r^.H- 71 . 1 3 B 7^.1 n ."I 71 ,G 3 A (,3-f ^3.V ^^-? column for the normals, and the right hand column for the retards. Note that in every case save one, grade 7B, the accelerates rank clearly above all others, and that in every case the retards rank lower than all others. This comes out with remarkable clearness when one examines a given age group throughout the city. This has been done for three age groups; first, those 13 to 13>^ years old; second, those 1214 to 13 years old; and third, those 10. to lOj^ years old. All the children 13 to 13^ years old were separated from their respective grade groups, and the average standings for these children were computed by grades. The other two age groups were treated sim- ilarly. There are children of each of the first two age groups to be found in all grades, 3A to 8B, while the 10 to 10^ year olds are found in grades 3.\ to 6A. Table 11 shows the standings bv grades for each of these three age groups in comparison with the average for the grade from which they are drawn. SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES 29 , ]< CO HH a, w ]7.x 73.1 3 A S'J.Z 31.1 HH.o S'S'.a. The distribution of 13-year-old children throughout all grades is striking evidence that chronological age means almost nothing in school organization. We discover that the children of a given age represent very wide differences in ability. Circles were drawn about the figures in the table which represent the scores of the normal children. Figures above these circles are the scores of the children who are accelerated, and those below the circles are for the children who are retarded. A brief study of this table reveals the fact, that what appears in general, as shown by diagram 8, .appears in particular for a given age group, viz., that the accelerated 13 to 13J4 year olds rank above the average for their grades, that the normal 13 to 13J/^ year olds rank above the average for their grades, and that the retarded 13 to 13J/2 year olds (6 months retards excepted) rank below the average for their grades, and increasingly so as they are more and more retarded. A comparison of each of the other age columns in table 11 with the "Average for Grade" column shows the result to be ex- actly the sapie. That is, the difference between 13 to 13J/^ year old children is much greater than the mere difference between grades SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES 31 3A and 8B, that is, 5 years, for while the 13 to 13yi year olds in 8B stand 3.9 per cent, higher than their grade, the 13 to 13^ year olds from grade 3A rank 28.1 per cent below the average for their grade. These things have been known to be true before, but here is exactly how true they are in Oakland, when examined from one single point of view in the curriculum. In making this general average for the city as a whole, and for the upper grades, Oakland has used children approximately a year older than should normally be used for these tests. The differences in age, amounting to from 6 to 8 years, which appear in every grade group, cannot be fully avoided, even under ideal conditions, but a better arrangement than the present would be to make the one year retards the normal gronp, calling the present normals one year accelerated, etc. See dia- gram 7. If the fifth grade children who made 100 per cent, in the test had taken the sixth grade test instead of that for the fifth grade, it is probable that many of those lOO's would have been reduced to 90,s, a few to 80's and, a very few to 70's. In spite of the fact that the accelerates have lifted the general average for the city very decidedly, yet the same is true of the normals, and also of those only one or two years retarded, as may be seen from a study of table 11. Again, if all the 13 to 13i/^ year olds in table 11 had taken a test one grade higher than they did, it is fair to assume that all the figures in that column of the table would have been lower than they now are. Just how much lower we cannot say, but the three averages in that column of the table which fall below 70% include but 41 of the 874 children included in the "13 to 13J/^ year olds" column. Which means that they might have been reduced quite decidedly without making the general average for children of that age lower than 70%, which is the desired standard. So, while it is true that it is the average child who has held down the general averages everywhere, yet they have not held it below the 70% mark. The only conclusion is that Oakland would have made an average of approximately 70% in the 4th, 5th, 6th, 7th, and 8th grades even if each of these grades had taken the test for the next grade above. This would not have been true for grades 2 and 3. Such facts as these are at the basis of administrative as well as teaching difficulties. First, from diagram 8, and tables 10 and 11, we note that in practically every case the accelerated child is ahead of his grade. This means that he is not having to work very hard to make a respectable showing. Exactly the reverse is true of the retarded child, and the more he is retarded the more true it is. If an intelligence test were applied to our 13 to 13yi year olds,, who by an age grade measure are accelerated, it is quite reasonable to suppose that in terms of their psychological ages they are actually 32 OAKLAND SPELLING INVESTIGATION retarded. And this simply because they have not had to work up to their full capacity. The opposite of this would likely be shown for the retards. This problem is being studied diligently, and in time we shall know the extent to which such facts exist. Meantime, by use of special rooms, for bright pupils as well as for dull ones, irregular promotions, and assistant teachers, the school machinery should make provision for teaching the bright children to work hard. To the teacher the demand is clear. She should be fully conscious, not only that this state of affairs exists in her room, but of the way in which these differences manifest themselves, and of the extent to which they exist as well. In her assignments she should plan to use up systematically that time which accelerates usually trifle away. Table 11 suggests strongly that there are quite a number of pupils in regular classes who ought, both for their own sakes and for the sake of other children, to be dealt with in rooms for subnormals. The amount of retardation for the city is too high. It carries no advantage whatever, but tends to discourage and depress. The year's work in spelling should be made reasonably difficult, but there is no reason for making it so difficult that only one child in three can complete it in the given time. It should be said, however, that the general results of this test do not seem to indicate that spelling is the cause of a very large per cent, of retardation. Which may suggest that spelling is receiving proportionately more than its share of time, for further discussion of which see chapter V SEX AND SPELLING EFFICIENCY. No one has yet argued, from a wide knowledge of facts, that sex differences show themselves so prominently in intellectual work as to demand that girls and boys be taught separately. This part bf the study was undertaken with a view to being able to state quan- titatively, what practical effect sex differences might have in a test of this sort, and to gain from the showing any suggestion it might have for the teacher and supervisor of spelling. The general findings are presented by grades in table 12 and in diagram 9, from which it appears that there is a clear difference in favor of the girls, that the difference is constant, and that it roughly increases with age. (Since so large a number of children are in- cluded, it may be of psychological interest that the increase in dif- ference is not more regular than it is.) These same facts are pre- sented for each grade separately by schools in diagram 10, which shows that grade groups vary in different schools, but that even when the boys are ahead they are rarely very far ahead. The fact that the difference is more pronounced in the upper grades is also clear. That the differences are much greater in individual schools than when the average is taken is to be expected. SPELLING EFFICIENCY AND INDIVIDUAL DIFFERENCES 33 TABLE i2. Grade Averages by Sexes, SEX GRADE 8 1 6 5 4 3 2 Bocfs 1Z.^ 77.0 78. ¥ S"/-? 7^.s: i>¥.9 ^<7,« Gi-y/s 18:3 8;2.¥ sav ss:^ 7LU ^f.^T ,6/. 9 Gi-r/s'^i/pen otrtti^ s-.y s.¥ JZ.0 3.S AS ^.4 ;2,7 V^ Just what this means for the teaching of spelling, in detail, may call for a careful study of these scores in relation to the relative ages of boys and girls, and also of the relative variability of standings for boys as compared with that for girls. For the purposes- of this DIAGRAM 9. Relative Standing of Boys and Girls by Grades in Forty Schools. i 1 (> f ^ 3 X. Qra. m fa & r. S a •S g s CO FQ Ml '^ V 'J! o IS ^ 4 *j O o X CJ CO >• < 111 »- m bl o o J3 of o o o t« a. •a ect to the amount of knowledge the child possesses, in respect to his methods of study, and in respect to the interest which he has in his work. It is the teacher's business to see that children are working up to the limit of their capacities, and not allowed to dawdle and wait for a dull or stupid classmate to catch up. This requires knowledge of individuals such as that suggested by these studies. The setting forth of the general features of such problems, and direction in how to solve them, are legitimate functions of the supervisor ; the con- stant watch to see that these ends are being met by teachers, is a le- gitimate function of the principal. Such assistance should be wel- comed, even sought after, by the teacher. The results of such co- operation will certainly place in clear light any new demands which problems of individual differences may make upon matters of or- ganization and management. In respect to sex differences, we note that they are definite and constant in their showing on the test. They roughly increase in significance from year to year, but with some irregularity, being least in grade 4 and greatest in grade 8. In amount these differences range from 1.8% to 5.7% by grade groups throughout the city ; but when grade groups in single schools are studied, we find these differences quite variable, often going as low as zero, not infrequently with the boys slightly ahead, and often as high as 10% in favor of the girls. These differences between boys and girls are very genuine, though in most cases not .so wide as to prevent their being handled in classes together. The main suggestion which these facts have for the conduct of the schools, is, that if girls mature more rapidly than boys, they should be provided with the chance to get through the elementary school at an earlier age. The amount of difference which appears and the way in which it persists, suggests that the 40 OAKLAND SPELLING INVESTIGATION girls should, on the average, finish the elementary school work in spelling, certainly a half year, if not a year sooner than the boys can finish it. It is not a case demanding rearrangement of school machinery, so much as it is one which demands attention to those more minute adjustments which the elasticity of the present machinery provides for, and which the teacher, supervisor, and principal, are respon- sible for effecting. Such educational capital as exists in these indi- vidual differences must not remain uninvested. They are the sav- ings accounts in school training which we ought to start early, for, however small, they are dependable. Our third trait, general intelligence, brought . to light further evidence that very wide differences in ability exist between members of the same class, that the teacher seldom had difficulty in nam- ing the brightest and dullest children in her room, and that spelling ability corresponded quite fully with this general ability. These differences are often wider than one would reasonably expect, between children a grade apart, and emphasize the fact that an assignment of work to such a class cannot avoid doing injustice to some of the members, either by making the lesson too easy or too difficult. Likewise the recitation period must waste tirne for some, permitting them to form slovenly habits of inattention, while others are struggling in the dark, more or less in a vain effort to comprehend. Such classes can be tested, but not taught. The effect of such conditions on the teacher has been pointed out above. Finally, then, the demand for "economy of time in the elementary school" can be effectively met by perfecting organization, and re- fining those inner and more minute adjustments for which teacher and supervisor are responsible, all in the light of the definite claims of individual differences which this test has brought to light, and stated in quantitative terms. CHAPTER IV. SPELLING EFFICIENCY AND SOCIAL DIFFERENCES. THE NATURE OF THE PROBLEM. The group of influences studied here are those produced by the child's contacts, through institutions, occupations, and race connec- tions, with the world outside. The purpose being to discover what, if any, are the differences in children for which such forces are re^ sponsible, assuming that any such differences, which would be of importance to the school, might manifest themselves to some de- gree in such a test as this. It is undoubtedly true that original endowment, or human nature, is extremely variable, and that it accounts for many of the important differences we find among school children. But we must keep in mind that it is just this original nature that education is trying to change, and that whatever plays a part in the stimulation and direc- tion of these native tendencies, is producing change in them, that is, educating the child. The school is but one of the many forces which are constantly playing upon child nature, and the school is presumed to represent systematized effort to produce certain well defined, and desirable changes. It is such only in proportion to the extent to which it corrects, supplements, and extends the influences of that larger school, life. The school must, in other words, have per- spective for its work. It must know that it stands beside the home, the church, the street, occupation, social activities, and ancestral in- fluences — which are quite as much social as biological, — and formu^ late its procedure in terms, not only of original human nature, but also in terms of both the shortcomings and the sound influences of these other agencies. Only by taking full stock of the work being done by these vari- ous outside influences, therefore, can the school shape the details of its own tasks intelligently. To the end that some little light might be thrown upon such in- fluences, the results of the test were studied from the standpoint of the father's occupation, the occupation the child desires to enter when he leaves school, the father's nationality, and the home lan- guage. Only records of selected groups, and not of all the children in the city were used in these studies. FATHER'S OCCUPATION AND SPELLING EFFICIENCY. For this study 2644 children from 13 schools, and from grades 2 to 8 inclusive were chosen. Not less than 200 nor more than 400 records were chosen from any one grade, and classes were taken from the largest as well as from the smallest schools, and from schools in all sections of the city. The group is a random selection therefore, and representative of what a larger number would show. 42 OAKLAND SPELLING INVESTIGATION The occupations of these 2644 fathers, representing 73 dififerent callings, have been thrown into groups as shown by the left hand column of Table IS". This grouping was used because it seems to divide people roughly along social lines, which are likely the result as well as the cause of their being in this or that occupation. The home of a professional man represents, on the whole, a higher standard of living than that of the clerk or labor group, and some- what different from, if not higher than any of the other groups. The number belonging to each of these groups of occupations is indicated in the right hand column of Table 13, which table repre- sents the complete distribution of the standings of the children of these fathers. Each of the groups is large enough to constitute a fair sample of the Oakland school population. The test averages for the different groups vary from 73.3% for TABLE 13. Showing the Distribution of Standings of 2644 Children by Occupational or Social Groups. Standings Ave. No. In- 100 90 80 70 60 50 40 30 20 10 ed Professions . — 70 111 20 132 15 80 86 63 119 22 143 17 78 116 71 105 18 122 13 75 116 36 101 10 82 10 38 84 19 61 8 54 11 31 46 24 52 4 30 8 25 40 14 34 4 23 6 11 9,^ 4 18 2 15 4 5 20 2 14 15 4 14 8 2 2 1' 8 4 s 75.7 72.6 79.8 77.2 7.37 78.7 72 3 303 Skilled Labor and 631 Officials 88 Business 622 Agriculture 84 Clerks 349 Labor ■. 567 the labor group to 79.8% for children of public officials, which is a difference of 7.5%. It is noticeable that children of skilled work- men spell little better than do the children of unskilled workmen, that children of clerks spell better than children of either profes- sional or business men, and that children whose fathers are public officials (including policemen) spell better than all others, when judged by averages. In Diagram 13 the standings of the children are distributed by groups in terms of the percentag'e of children receiving a given rank. SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 43 '^OOf o o 00 CO O 2 & u ^H T3 ft cn *^ 9 o The full length of each bar represents 100% of the children of the group, each separate division of the bar representing the per cent, of the group receiving a given score. By reference to the scale at the top, and to the explanation, it is easy to see that of those re- U OAKLAND SMLLIN6 INVESTIGATION ceiving 100% standing, the professional group ranks highest, and the lafcr group lowest; of those receiving 90% or above, the officials group stands first, and the professional group fourth, with labor lowest. Examinations of the other sections of the diagram places the children of officials clearly ahead, with the professional group second, the clerks third, and business fourth, which is in part a differeiit order than when the groups were compared by averages. Whatever other factors may enter into the determination of these standings, such, for instance, as the fact that the labor group is heavily weighted with ignorant foreigners, they do not change the practical meaning of the showing in the least, for labor is so weight- ed in most of our American cities. The first fact is that there are differences. Before the school undertakes to do anything about these differences it must know why they exist. If the reason is something the school cannot control, then its duty is to try, not to remove the cause, but to cure the ill. The foreigner is here, and his children are becoming Americans. The laborer is one of our company, and though his children may be on the average of lower mentality than children from the professional group, we must ask does the school know for sure that these children are in school as many days in the year, that they are not more often sick or physic- ally defective, that they have equal school opportunity with children from other groups, and that they are equally able to avail them- selves of those opportunities? It is this end of the situation which the school can study, and, to a large extent, control. This study is further evidence that economy of time for the school may lie in the direction of economizing the time and school privileges for individual children. Variety in home life from the economic standpoint points to another set of individual differences which school machinery and teaching must recognize. CHILDREN'S OCCUPATIONAL AMBITIONS AND SPELLING EFFICIENCY.! In this study are included 3079 children, or practically all the classes in grades 7 and 8 throughout the city. The question here is, first, what occupation does the child wish to enter, and second, how well can he spell, or is there any apparent correlation between the child's spelling ability and his present occupational ambition ? The total number of occupations mentioned by the children in- cludes 93, 30 more than the number actually followed by the fathers of the 2644 children studied above. There were 13 of these which were mentioned but once each, and these include butcher, baker, plas- terer, and reporter. On the other hand there were 14 occupations, each of which was chosen by 50 or more children. A study of the occupational choices of these children is itself an interesting ques- iSears, J. B. — Occupations of Fathers, and Occupational Chjoices o{ 1039 Boys in Grades Seven and Eight of the Oakland Schools. School and Society, May 22, 1915. SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 45 DIAGRAM 13. Distribution of 7th and 8th Grade Children With Respect to Their Choice of Occupations. 3079 in All. tion, but for present purposes Diagram 13 shows the distribution of their choices which is useful in this connection. One cannot help being struck with the fact that — according to figures in the above section of this chapter, — out' of 2644 fathers in the city, less than 13% are engaged in professions, while 45.8% of these 3079 chil- dren desire to enter professions. Some reconstruction of these am- bitions is ultimately necessary, and the school should not shirk its very evident responsibility in this connection. This is incidental, however, to the question of what relation there may be between these wise or unwise ambitions, and the ability of the children to spell. It should be pointed out in passing, that there is little ten- dency for the son to select the pursuit his father is engaged in, con- sequently bookkeepers here are by no means the sons of bookkeep- ers, engineers are not miainly the sons of engineers, etc. Here the ambitious son or daughter of a laborer may, and very often does, appear in the professional, or business group, and vice versa. From the above diagram it may be seen that teaching and en- gineering rank very high among the professions chosen, and that 46 OAKLAND SPELLING INVESTIGATION TABLE 14. Occupational Distribution of the 7th and 8th Grade Children and Their Standings by Groups. Occupation No. Children Choosing Average Standing of Group Professions : Teaching 346 362 17 684 1409 485 72 557 318 87 30 38 598 24 18 80.2 Engineering 76.9 Writers 81.2 Others 78.0 78.3 Business : Business service 79 3 Others 76 1 78.9 Trades 74 3 Agriculture Labor 71.5 73 Housewife 82 1 No choice 78 3 Baseball Others 62.5 73 3 Total ; - 3079 77.7 business service, by which is meant stenographers, clerks, book- keepers, etc., includes most of the business group. The number of children choosing these different lines, and the group averages they make in the test are shown in Table 14. The average for all is 77.7%. For the professional group of 1409 chil- dren the average made is 78.3%, while the business group of 557 made 78.9%. Those expressing no chioce rank even with the pro- fessional group, and the trades group, 318 in number, drop to 74.3%. The small group of 38 who wish to be house wives ranks above all, making 83.1%, while the two dozen who wish to play baseball rank far below, with an average of 62.5%. In interpreting these figures we need to keep in mind the fact brought out by Table No. 13 above, that girls in grades 7 and 8 spell on an average from 5.4% to 5.7% better than boys. The group of teachers, who rank high are largely from among the girls, and so the engineer group, all of whom are boys, cannot be counted specially low when they fall 3.3% below the teachers. They are low, however, as compared with the professional group as a whole. A more interesting comparison of groups is seen in the following arrangement : SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 47 Chosen mainly or wholly by Chosen mainly or wholly by boys. girls. Average Average Engineers 76.9 Housewives 83.1 Trades 74.3 Musicians 81.1 Labor 73.0 Stenographers 80.5 Agriculture 71.5 Teaching 80.8 Base Ball 62.5 Nurses 77.6 Bookkeepers 77.8 By a comparison of the groups on the left with those on the right, we find that sex does not explain all the differences. For in- stance the difference between the average for the agriculture group, and any group on the right, is greater than the normal difference due to sex. That is, agriculture and baseball have attracted a much lower type of spelling ability than that attracted by music or nursing. It is interesting to note that all the groups on the left rank below the lowest group on the right, and that bookkeeping, which attracts both boys and girls, ranks approximately between the two. Even if sex explained these differences, which it does not, we have still to consider those differences which exist between the engineer and baseball groups, or between the housewife and nurse groups. Clearly baseball does not attract brains, if spelling efficiency meas- ures brains. And it is the lower intelligence class among girls who wish to enter nursing, while music and motherhood call out the best minds there are among the girls. So, measuring boys against boys, or girls against girls, or boys and girls against each other, we find that lines which divide them with respect to their occupational ambitions, divide them also with respect to their spelling efficiency. It may not be that low spelling efficiency causes a boy to wish to become a ball player, nor vice versa. There may be other things: race, health conditions, low intelligence, home life, etc., which enter in, but it remains true that, out of a total of 3079 children, the 34 boys who desire to become b^ll players are notoriously poor spellers, that the 196 girls who wish to be musicians and the 295 who wish to become stenographers are decidedly good spellers. So, whether general intelligence, or special spelling intelligence, accounts for these occupational choices, does not appear in this showing; yet it does appear by these facts, that the occupational choice of a child indicates roughly what his relative spelling ef- ficiency is. If the pupil aspires to be a teacher, or a writer, or a musician, then the chances are that in spelling efficiency that child ranks relatively high. Or if the ambition looks toward baseball, or labor, or nursing, the chances are that in spelling efficiency the child would rank low. 48 OAKLAND SPELLING INVESTIGATION Now, to try to state exactly what is the extent of these differences in general intelligence is not necessary and probably not safe on the basis of a single limited study of one kind of intelligence. The differences are here, and if spelling efficiency and general intelli- gence go together, as our showing in Chapter 3 at least suggests, then a boy's standing in school work indicates roughly the direction his occupational interest is likely to take. This is clearly suggested below. Did the good spellers select occupations in which good spelling is an important qualification ? Stenographers made 80.5% ; teachers, 80.2% ; writers, 81.2% ; bookkeepers, 77.8% ; bankers, 81.5% ; lawyers, 78.6%; librarians, 81.7%; secretaries, 80.0%; office work, 74.4%. These would all need to be good spellers, and all save the last group rank above the average for the total number. On the other hand we find that the miners made 75.0% ; farmers, 71.7%; actresses, 76.5%; artists, 74.9%; architects, 74.0%; carpen- ters, 69.4%; draftsmen, 69.4%; mechanics, 72.8%; seamen, 73.3^'' ; druggists, 68.5%; .doctors, 77.6%; dressmaking, 73.0%; millinery, 79.6% ; and salesmen, 65.3%. Only the milliners rank up to the total group average. Most of these need to know how to spell, but not to the same extent as do those of the group above. Roughly, then, those who have chosen an occupation which de- mands unusual spelling ability are the best spellers, while those who have chosen occupations in which ability to spell is not of first im- portance, are below average in spelling efficiency. There are some exceptions, as, wireless operators made only 73.5% ; mail clerks, 70.0% ; merchants, 66.7% ; while actors made 81.2% ; auto repairers, 78.6% ; and aviators, 80.%.; and hairdressers, 84.0%. Of course we are dealing with small groups in making these statements, and their significance lies not in the fact that it is true that 6 librarians, or some other small groups, average high, but that nearly all of these small groups who ought to be good spellers, are, and nearly all those groups who need not be good spellers are not, and the exceptions are not numerous. There are a few rather striking facts, which at first appear diffi- cult to explain, such as in the difference between the standings for the 131 electrical and those for the 125 civil engineers. The former average 73.9% and the latter, 80.6%. Again, 8 actors average 81.2%o, while 23 actresses average but 76.5%,. The 16 high school teachers average 85.0%, and the 95 teachers of manual training, domestic science, etc., average 78.6%, and the 227 grammar school teachers average 80.4%. We are likely to say that electricity and manual and domestic work appeal to certain students because such studies seem to offer relief from books, and that these fields will attract more of the average intelligence group than do the usual lines of teaching or the less exciting lines of engineering. Hence SPELLING EFFierfiNCY AND SOCIAt DIFFERENCES 49 such facts as these do not mbdify the 'general statement of the re- sults which this study shows, viz: that the type of child which aspires to enter a calling which will demand superior spelling abil- ity is a child who possesses such ability, whereas the child who has not spelling ability will usually not aspire to a calling which will demand such ability. FATHER'S NATIONALITY AND SPELLING EFFICIENCY. For this section the records of practically all the 3rd, 5th, and 8th grade children were studied. These give us typical groups, in- cluding the grade which received the poorest standing in the test and the one which received the highest standing, also the youngest children, a middle age group, and the oldest children. That we may know the distribution of children by the nationali- ties of their fathers the showing for grades 3 and 8 are pre- sented in Diagram 14. The distribution of grade 5 would doubtless fall somewhere between these two, though the per cent, of foreigners is much lower here than in either the 3rd or the 8th grade. A little study of these two grade groups is of interest in corhparing the showing for grade 3 in the test with that for grade 8, either as is done in this section, or as was done in Chapter 2 above. First of all, grade 8 appears to have nearly 2}^% more native children than has grade 3. Second, of the foreign groups the British have more than doubled, the Ger- manic group has increased nearly one-third, and the Scandinavians have more than held their own, while the Romanic group has de- creased by approximately two-thirds of its number. Of the British the chief gain is with the English and Scotch, the Irish showing a slight decrease. Among the Germanic groups the chief gain is with the German and Dutch. With the Romanic groups the Spanish has disappeared, the Italians and Portuguese have dropped off more than two-thirds of their number, while the French has gained slightly. The composition of the group has therefore changed from slight- ly over one-half to close to three-fourths English speaking in passing from the 3rd to the 8th grade of school. Why this percentage of American fathers is so much higher for the 5th grade and then falls back again does not appear. Table 15 presents by per cents, the entire distribution of the .standings made by the children of the three grades, first for the grade as a whole, and then for the children of foreign fathers from each grade. On the right of this table is the average for each grade and foreign group. Taking those making 100% in the test, it will be seen that 20.4% of the children of grade 8, that 33.4%. of those of grade 5, and that 13.6% of those of grade 3 are included. While for the 50 OAKLAND SPELLING INVESTIGATION SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 51 TABLE 15. Showing the Per Cent of Children of Native and of Foreign Parents Who Receive the Different Standings and for Those of Foreign Parents Separately. Children included Standings Ave. Stand- ing 100 1 90 80 70 60 50 40 30 1 20 10 All of Grade VIII 20.4 19.0 18.7 18.9 16.5 16.8 15.3 15.6 13.9 13.5 11.4 12.1 13.5 13.6 10.1 12.3 5.5 6.6 9.8 10.2 8.4 9.0 3.7 3.9 9.1 9.8 4.1 4.5 2.0 1.6 7.7 7.9 3.7 3.4 1.3 1.5 5.7 5.9 1.3 1.2 .4 .7 4.0 4.3 .3 .5 .2 .7 2.7 3.8 : .3 1.8 2.8 76.3 Foreign in Grade VIII All of Grade V 17.6 33.4 30.9 12.6 9:2 18.9 25.5 25.9 17.8 16.9 74.3 84.7 Foreign in Grade V All of Grade III . . 82.4 66.7 Foreign in Grade III 63.6 foreign group we find approximately 3%i less in each case. Of those receiving 90% in the test we find the foreign groups slightly- ahead in grade 5, almost equal to the grade average in grade 8, and nearly 1% behind in grade 3. For the 80% standings the foreigners are slightly ahead in each case. With slight variation the difference in favor of the foreign groups increases as we read toward the lower standings. By the averages shown at the right of this table it will be seen that children of native and foreign fathers together spell 2% better in the 8th grade, 3.3% better in the 5th grade, and 3% better in the 3rd grade than do the children of foreign fathers. The distribution of these standings may be more easily seen, and the groups moreeasily compared by reference to Diagram 15. Here it is easily seen that the 3rd grade, more than the others, suffers from the presence of the foreign groups. We will recall that the per cent, foreign is larger in the 3rd than in the 8th grade, and that the Romanic groups are relatively large in the 3rd grade, and that the British groups are relatively large in the 8th grade. In order to see more clearly the effect of this difference in the composition of the foreign groups in these two grades, they were studied by nationalities, and their averages, so grouped, were compared with the averages of their respective grades. '52 OAKLAND SPELLING INVESTIGATION a a in -H •IH 13 a «0 I— I ^ H is t— I a- uj ^ ^ &« ^ ^« Qi Ki C£> "o ><^ a. w s as Q- o LlJ ^ < t< or «> D 2 w////m/////j ini/iiinniiiiniiTm \7777m/////uiniiinnnm]m ^/)/////////////////////////////////W/w\ 5 T^^:^^^ k WWWWW 'Km\\m\ 'K\\\\\m\\\v IWWWWWWWNM ^ MSSS^SS^ kv^vmmmmw^ 01 qz oq. 0]r og 09 OL OS 01 r m\\mm °^ ^5 SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 53 bc a u IB a g ^ o Pi o go 'S o o I aeananj-Hod ^8 ? 5§ These comparisons are shown in Diagram 16. In each case the grade average is shown by the straight line, with which the averages for the different race groups may be compared. It shows the Portu- guese low in both cases. But we have seen that very few Portu- 54 OAKLAND SPELLING INVESTIGATION guese children remain in school throughout the eight grades, hence they do not lower the average for the 8th grade very much. The few Italians who remain through to the 8th grade are superior pupils, but the larger mass of them who start in in the earlier years fall far below the grade average. The Germanic and British groups which we should ordinarily expect to rank well up to the grade average are in all cases below it in both grades, with the exception of the Austrians, who rank high in grade 3 and very low in grade 8. The Scandinavians vary, the Danes and Norwegians in both cases being low, and the Swedes in both cases high. It is the Swedes, Austrians, and Russians who keep the 3rd grade foreign average up, while these plus the Italians keep the 8th grade foreign average up. Of all the foreign groups, only the French and the few odd Oriental races, grouped as "all others," fail to do better in the 8th than in the 3rd grade. Which .means, of course, that it is the better class of foreign children who remain in school through all grades. It must be remembered that children of foreigners are not measured against children of natives alone here, but against the entire grade, which includes themselves. The difference of 2% to 3.1%? therefore, is not the full extent of the difference between children of natives and children of foreigners. A school here which is mainly foreign would likely rank not 2% to 3.1% lower than the city average, but more likely about 5% lower. INFLUENCE OF HOME LANGUAGE ON SPELLING EFFICIENCY. Teachers whose classes are made up of children from homes of foreigners frequently complain that they cannot maintain high standards because their pupils are all foreign children. As to whether this is true or not there is little doubt, but the ways in which, and the extent to which it is true, are important questions. If the foreign child has greater difficulty in getting on in school, is his own mother tongue a handicap? If so, how, and to what extent? If the child hears and speaks no lannguage at home except Italian, or German, do the Italian and German speech habits inter-- fere in the child's attempt to use English ? Or does this lack of drill in the use of English at home affect his work in school? Under this general question comes our question here : does this hearing and using a foreign language at home show itself in the types of errors the child will make in spelling? Is there, on the one hand, some- thing characteristically German, or Italian, or French, in his at- tempts to spell, or is there' merely a poverty of knowledge about English, which shows itself in a poorer variety of guesses at how to si>ell, or is it both, or neither of these? A certain learned man insists that he became a poor speller of English after- having studied French.' If French disturbed his English spelling, the opposite may easily be true of foreign children learning English as a foreign language. SPELLING EFFICIENCY AND SOCIAL DIFFERENCES . 55 By the time a child reaches the 8th grade his English habits must be well established. If any of the influences suggested above are of consequence therefore, it would be in the early grades.- The errors of 3435 4th grade children were used for this study, and classified according to the language spoken in the home by the father and mother. This made 40 different classes of errors for where English and a foreign language were both used that was made a separate group. The number of children from homes in which English -was riot spoken at all is 474, or about 30%. In- cluding- those in which both English and a foreign language are usedttiere are 535, or about 23% of all the children of the grade. Of these foreign languages Portuguese and Italian are the most im- portant and are the only ones treated separately below. The relative number of errors made by children of foreign and of native parentage is in effect discussed above, but the relative num- ber of different kinds of errors was not mentioned. Table 16 shows these facts for the children of grade 4 by words. In reading this TABLE 16. Comparisons of Errors Made by Children from English Speaking Homes With Those Made by Children from Foreign Homes. Word Number of different ways misspelled by No Portu- Italians All other Total English All to- attempt guese Foreign Foreign gether made forty 12 13 35 34 54 71 17 rate 17 13 38 51 95 120 26 children 19 12 23 43 99 138 31 prison 22 14 40 58 89 116 19 title 28 24 41 73 117 163 60 getting 6 10 15 25 34 49 5 need 17 12 19 31 52 69 2 throw 14 10 19 27 66 80 24 feel 17 17 29 42 78 91 30 speak 9 7 10 18 41 48 22 table we need to keep in mind that there are four children from English speaking homes where one comes from a home where a foreign language is all that is heard. " in mind that there are four children from English speaking homes where one comes from a home where a foreign language is all that is heard. First, taking the word "forty," we note that 13 errors were made by children from Portuguese homes, 13 by children from Italian, and 35 by children from other foreign homes, in all 34 dis- 55 OAKLAND SPELLING INVESTIGATION tinct errors were made by children from foreign homes, while- 54 were made by children from English speaking hornes. Thgt is, the fareign children found 34 distinct ways and the English 'children found 54 distinct ways to misspell the word "forty." Together they found 73 distinct ways. Assuming that these errors are mere guesses, for all but three or four of them are such, as will be seen later, then it appears that 535 foreign children made 34 guesses, while 1780 English children made only 54. But since there is a limit to the number of errors possible, or at least reasonably prob- able, it is not possible to say which has made relatively the greater variety of guesses. This can be said however, that of the 34 differ- ent errors made by the. foreign group, 17 of them were made by one or more English speaking pupils as well, while of the 54 made by the English speaking group, there were 37 which were not made by any of the foreign group. Without giving the figures, it may be said that the errors for the other words run about this same way. If there were not some limit to the number of diflferent errors pos- sible, then it would appear that the foreign children furnish a rela- tively wider variety of guesses than do the English speaking children. But there is such a limit, and so we should not expect the number of different spellings made by the English speak- ing children to be 4 times the number made by the foreign group. In the variety of errors therefore, it does not appear that the child from a home where no English is spoken is at a special disadvantage. We must also remember here that the children of foreign parents do not spell as well as do the children of native parents, (see above). Secondly, what is the distribution of these errors for the foreign as compared with the English speaking group? This is revealed by table 17, which shows the three or four most popular spellings for each word as they were made by the different groups of pupils. It will be seen that an error which is popular with one group is usually just as popular with the other groups. Table 16 shows that the total number of errors for the 10 words range from 49 on the word "getting" to 163 on the word "title." All spellings made by at least 1% of any group are included in Table 17. A comparison of the columns of this Table shows the relative popularity of a given error with the different groups. In this Table the correct spelling of each word is followed by the chief errors. The vast number of errors not shown in the table merely means that most of the errors are made by only one or two children, sometimes by a foreign, some- times by an English speaking child, and very frequently by both. Referring to the word "forty" in the table, it will be seen that 70.8% of the English, 64.3% of the Portuguese, 61.1%, of the Italian, and 69.5% of all other foreign speaking children spell the word correctly. The most important error, f-o-u-r-t-y, is made by 21.3% of the English, 24.6% of the Portuguese, 27.6%. of the Italian, and 30.8% of all other foreign speaking children. The SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 57. TABLE 17. Comparative Distribution of Spelling Among Different Nationalities. How the Word Was Spelled and by What Per Cent of Each Group. All Other Word Portuguese Italian Foreign English forty 64.3 fourty 24,6 forthy 1.3 fourthy 1.3 rate 82.8 rait 3.4 raite 0. wrate 1.8 children 84.9 chrildren 2.7 childern .68 childen .68 prison 80.8 prisen 2.7 prision .68 title 55.5 tittle 16.4 titel 2.7 tital 1.3 tidle 1.8 getting 76.0+ geting 20.5 geating 0. need 60.2 nead 16.4 knead 8.9 kneed 2.05 neid .68 throw 61.6 through 25.3 thow , 0. though .68 feel _ 54.1 feal 9.5 field 19.1 feild 2.7 fiel 1.3 speak 91.0+ speek 3.4 speake 0. 61.1 69.5 70.8 27.6 20.8 21.3 0. 1.6 1.2 .74 .56 .55 76.8 80.2 79.6 9.7 4.3 3.4 . 1.4 .82 1.7 .74 .27 1.4 88.0+ 83.9 88.0 + 2.2 1.9 1.4 1.4 1.3 1.2 0. 1.0+ .67 85.0+ 84:9 86.5 0. 1.6 2.5 2.2 0. 1.2 66.4 61.6 66.1 14.1 11.2 9.8 1.4 2.1 2.8 0. 2.7 2.7 1.4 1.9 1.7 67.0 + 69.5 77.0+ 26.1 22.4 20.0+ 0. .56 .20+ 71.6 69.3 68.0+ 17.1 20.8 20.3 3.7 1.9 2.8 0. 1.6 2.1 0. .56 .68 61.6 63.8 60.2 30.5 23.0+ 27.7 .74 1.91 1.06 0. 1.0+ 1.4 51.4 65.2 59.4 11.1 9.5 13.6 14.9 7.9 8.03 .74 2.4 2.6 1.4 1.91 2.4 91.0+ 92.0+ 86.8 4.4 4.1 4.6 .74 .27 .73 58 OAKLAND SPELLING INVESTIGATION next most important error is made by less than 2% of any group and the third most important error is made by less than 1% of any group. A similar examination of the showing for the other words convinces one that children from foreign homes make the same errors as do children from homes where the mother tongue is English, and furthermore, that they make them in approximately the same proportion. Where an error is made by not more than one or two children out of so many it is clear that the spelling was only a random guess. To see what proportion of all spelling errors are due to random guessing refer again to the word "forty" in Table 17. First assume that f-o-u-r-t-y, f-o-r-t-h-y, and f-o-u-r-t-h-y, are specially attractive spellings for children to use on this word, that since so many chose these ways there must be something reasonable about them, then we have left 67 other distinct spellings to account for. But on adding the figures in our table for the word "forty" we see that these 67 ways comprise only from approximately 7% to 12% of the children. If from 20.8% to 27.6% of these children all together make only three types of errors on the one hand, and from 7% to 12% make 67 different types of errors on the other, then this seems to argue that these three errors are important and represent attract- ive lines of incorrect procedure for the child, whereas any one of the 67 others might never be selected since here it has attracted but one or two children. What we are interested in here is, that children from foreign homes do precisely what other children do and in about the same proportion. As was pointed out above, children from foreign homes do make errors which are not made by children whose mother tongue is English, but in all cases such errors are made by not more than two or three children, and so by that fact prove themselves not to be errors which are for some reason specially attractive to children of a given nationality, who hear only Italian, or German, or French at home. They are mere random guesses in which the order of letters nowhere seems to have been influenced by habits of speaking, hear- ing, reading, or writing the foreign home language. Wherever a word does look slightly French, or German, or Italian, it is fre- quently found that a native child produced the same combination. So, from this study it appears that the mother tongue, when foreign, offers no handicap due in a peculiar way to the home language. As was shown above, children from foreign homes do not spell' as well as do children of natives, but there is here no spe- cific evidence that their home language is responsible for this. The evidence that the difficulty is not in the effect of the home language lies in the fact that there, is no noticeable differance — save the slight difference in extent found above — between the errors made by SPELLING EFFICIENCY AND SOCIAL DIFFERENCES 59 children from homes where only Eng'lish is spoken and those where only a foreign language is spoken. No effort is made here to classify these errors. Some of those in table 17 are easily accounted for on the basis of incorrect hearing. But whatever affected the native child affected similarly the child of the foreigner, hence we are not concerned with any other sort of analysis of the errors here. One word may be said however, viz. that when we know that there is one chance in ten that a child will spell "title" t-i-t-t-1-e, then we know one very important thing to fortify the child against when we teach that word. This is just as true for the foreign as for the native child. And a second thing we know is, that there is a fair chance that a child, English as well as foreign, will simply guess at his spelling. We should also fortify with strong habits, dictionary habits in the upper grades, against all guessing. SUMMARY AND CONCLUSIONS. In this chapter it is not meant to assume that the standings made by the children studied are due exclusively to the social fac- tors considered. These factors are no more cause than they are effect. Just as different occupations attract different individuals so different individuals select different occupations. Technical occu- pations do not attract people who haven't intelligence enough to enter them, nor do men of "unusual ability usually choose to be hod- carriers. The school wishes to know the material it has to work with. It is as much concerned with differences as with likenesses, because its problem is always with the individual. If the child's contact with the institutional life brings out individual differences which the school ought to be cognizant of, then the school ought to study children with respect to their institutional relationships. Our study of 2644 children grouped by their fathers' occupa- tions show that children of laborers rank lower in this test than do the children of any of the other groups. The differences between these groups, most of which are not large, may be due to original en- dowment, or to training, or to both. The point is, they exist, and m^ust be due to something. The school tends to make children uni- form in their studies and that means that the school obliterates just such differences as these. Our question is, does it do so by holding the brighter groups down to the speed of the duller groups, or by ac- cepting a lower grade of work from the slower children? When the school knows that differences exist, and then knows where those differences are, and how they manifest themselves, its problem then becomes specific in place of general and vague. Such differences are found to exist between the children whose parents are in different occupational groups, between childrn in different 60 OAKLAND SPELLING INVESTIGATION groups according to their occupational ambitions, and between children of native and foreign parents. Not only do we know these dififerences exist with respect to spelling efficiency, but how great they are. The responsibility of the school lies in the direction of refining the more minute adjustments of the school to the child, and seeing to it that every child is able and does enjoy his full educational op- portunity. The first move is to study the health and attendance records of the children, so that there can be no doubt that the child of the laborer, or the child who wishes to become a nurse or baseball player, or the child of foreign parents is as physically fit, and as infrequently absent, and as adequately equipped with school necessi- ties as are other children. When the school has done this and re- moved any existing differences of this sort it must accept any further dififerences as hindrances to proper grading, classification, and group teaching, and must meet them with special provision of separate rooms, assistant teachers, etc., and not permit the slow, work of a few pupils to retard the progress of the bright children who ought by all means to be kept busy up to their full capacity for work. All the results shown in this chapter save those in the last sec- tion, show definite differences to exist between children along cer- tain lines. Not only does economy of time in handling these chil- dren, but also economy of educational opportunity for the children as well, rest upon knowing what these differences are, their extent, and how they manifest themselves in their school work, and then upon refining the grading, teaching, and grouping of these children to the end that these differences shall not operate as retarding in- fluences upon class progress. This requires careful study by prin- cipal, supervisor, and teacher, and constant readjustment. It is not a question that can be permanently settled. Its solution is ever in process, and requires continuous alertness on the part of teacheir, supervisor, and principal. The influence of the home language seems not to be very evi- dent, since the errors made by children of foreign homes are in the main identical with those made by children whose home language is English, and are made in approximately similar proportions. Neither do the errors show by the order of letters any influence of a foreign language. That is, the difference found above between the spelling of children of native and foreign parents seems not to be due to the influence of the foreign mother tongue so far as the number and type of spelling errors could reveal this. CHAPTER V. THE ADMINISTRATION AND TEACHING OF SPELLING. THE PROBLEM. The administration of the subject of spelling has only recently come to be based upon something besides tradition. The erroneous idea that efficiency in spelling consists in being able to spell several thousand uncommon words, and that good teaching of spelling consists in a bare assignment of a long list of words, and later in testing orally or in writing to see whether or not those words have been learned, have in recent years become archaic iiotions, replaced by the idea that the actual number of words which a child should be able to spell is relatively small, and to be determined by his written, not his speaking and reading vocabularies. In like manner psychol- ogy has shown the futility of undirected study, of bare drill, as a method of work for the child. From these studies of psychological and practical facts, and to some extent from the more modern handling of the cost side of public education, and the demands of the business world, have arisen problems in the administration of spelling, such as : what shall constitute the subject matter of the course, what place shall it occupy in the curriculum and in the daily time schedule, what its relation to other subjects, the grading of the words, the length of lesson assignments, etc. ? In other words, the modem superintendent has an educational policy which is just as well defined and as con- structive in respect to spelling and its function in the training of children as it is in respect to the construction of buildings or the purchase of equipment. Such a policy sees spelling not as a sepa- rate subject alone, but in its relation to other subjects, as a part of a child's total equipment. The tendency today is not merely to formulate such a policy and leave it to its fate in the hands of principals, busy with admin- istrative details, but to secure highly trained officers whose special task it is to put that policy into practice by painstaking supervision of the work of teachers. Oakland is not indefinite with respect to her policy as it affects spelling, nor is she without competent officials through whom to administer that policy. In addition to the printed course of study the writer is indebted to Miss Ida Vandergaw, supervisor of pri- mary work, for a detailed statement of the policy affecting spelling, which policy has been in effect during the past five years in the city's schools. Stated in few words, and without the pedagogical and psychological reasons set forth in the communication, the plan is as follows : OAKLAND'S POLICY AFFECTING SPELLING. During the first year spelling is not taught till the last six weeks. During this year composition is "oral only," and writing 62 OAKLAND SPELLING INVESTIGATION "at the blackboard entirely during the first term, mainly there dur- ing the second. Seat work, large free exercises." Reading is taught by a phonetic method and consequently must not be con- fused by the study of spelling, which deals with letters and their order instead of with sound elements. In the second year the spelling work aims to develop the habit of noting words as made up of parts : e. g., fishing f-ish-ing. Also the "haJ^tpf never gues§i«g" at the spelling of a word. The words studied are "selected mainly from words- asked for during com- position lessons." The writing continues as blackboard work, seat work with pencil, with "written spelling and composition on the blackboard." Written composition begins in the second half of the year, and so creates a demand for knowledge of spelling and writing. The justifications for this policy are clearly stated, and are from the standpoint of psychology and physiology sound. The teaching of reading is the first business of the schools in Oakland, and all other formal work awaits progress in this subject to a certain point. There are certain important questions to be answered: first, how long will it take to establish control of the mechanics of read- ing to the point where learning and using the letters in spelling will not interfere with the handling of words from a phonetic ap- proach? Oakland's answer is six weeks less than one school year. Second, How old must a child be, or how long must he have been in school, and what training is needed before he can wisely under- take the physical and mental task of writing? Oakland starts the child to writing during the first year, but requires no written lan- guage work before the second half of the second year.. That is, a year and a half are spent on some kind of writing exercises before the child is asked to make writing a mode of expressing his thought. Third, How much drill in oral expression should precede work in written expression? Oakland says one and one-half years. Now, from the experimental knowledge we have, the order of approach, and the reasons offered for that order, are entirely sound. The question of speed is not yet so fully settled. It is one thing to say that the procedure should be first in terms of the coarser ad- justments, and quite a different thing to say how fast training in the coarser adjustments shall take place, and how far it must have proceeded before training in the finer adjustments can wisely begin. The difficulty lies, not in stating the general principles, but in stating the rate at which the different steps in the training may succeed each other. Oakland demands serious work on the ordinary penman- ship controls at the age of seven and one-half years. That is, pen- manship as a mode of expression has waited for this length of time the development of finer muscular and visual co-ordinations on the physical side, and the development of facility in oral expression and a legitimate motive on the intellectual side, while spelling has awaited the development of some writing controls, of facility in oral THE ADMINISTRATION AND TEACHING OF SPELLING 63 expression, and knowledge and skill in the phonetic approach to reading. All this is good order, and not much by way of achieve- ment in writing and spelling during the first two years is expected. In the light of this test that for spelling is lower than has been at- tained in other cities. Contrary to this plan of deferring the teaching of penmanship till there is occasion to use it for expression purposes, and conse- quently of spelling until need through writing exists for it, we see the Montesorri system beginning both of these subjects during kin- dergarten age, carefully creating motive, and getting results. But aside from this as an extreme, there is no doubt that the traditional practice in American cities has been to undertake both formal spell- ing and formal writing before the end of the first year. That is, Oakland has definitely and purposely departed from the traditional practice in this particular, and has thus far justified the procedure on theoretical grounds, and upon the assumption that superior work in oral expression has been accomplished. This policy has been in operation for five years, and, as shown in chapter 2, the schools stand very low in spelling in grades 2 and 3, possibly as a result. This must be offset by gain somewhere, else an added spelling burden is unnecessarily being placed upon the grades above these. As stated in chapter 2, mere opinion is not an absolute guarantee of this, and steps should be taken to make quan- titative statement of this supposed gain in oral expression by use of such measures as the Thorndike and Courtis reading and word test. As stated above, no fault can be found with the order of pro- cedure followed in Oakland (assuming that the policy is really be-, ing carried out), but until it is shown that the supposed gain in oral language and dramatics is a real one, the policy cannot fully recom- mend itself to the more conservative school public in point of time for beginning formal school work in writing and spelling. Not having looked for evidence one way or the other, the writer has no reason for supposing that such gain is not an actual one, but if so, it is all the more important that we should know about it. It is true that the sooner penmanship is mastered the sooner a child will make use of it as- a mode of expression, and so necessitate work in spelling. This must not go so far forward in the course as to con- fuse the child in his work in phonetics, nor in such intensity as to become hygienically objectionable. On the other hand, just because these subjects are formal, they need not be taught as abstract ma- terial, and so on that account deferred till the child is more mature. If the time for beginning the formal subjects depends upon ma- turity we must remember, first, that chronological age does not fairly indicate intellectual maturity, that every entering class, accord- ing to the Binet tests, contains children both older and younger mentally than six years. That is, in practice, maturity is not after all made the basis for classifying children as they enter school. 64 OAKLAND SPELLING INVESTIGATION School practice will make use of mental maturity as a basis for classifying children when a rrieans for determining maturity is per- fected, btft till then, we must- not set too much store by the as- sumption that i given child is too young for certain work because he is but six years old. While we must defer formal school work till it can be properly motivated, we must not fail also to look into the possibilities of developing motive as early as possible. In other words,' when to begin this or that study is as yet largely a matter of opinion, and only controlled practice will give us final answer. It is noteworthy, that cities are making their procedure more and more definite, and it is to be hoped that what Oakland has done by way of a slight break, not entirely exceptional, to be sure, with the traditional procedure, may help to solve .the question. We know that the anticipated effect of this on spelling has been realized, but we do not know the extent to which the anticipated effect on oral English, dramatics, etc., has resulted, and this Oakland is obligated to inquire into. It is the writer's opinion that when the results are known, the policy will in general commend itself. Its enforcement will require more careful supervision, for the reason that not only are results in content work for these years more diffi- cult to measure and check up, but because the material to be taught is less definite, and the methods demand a higher degree of teaching skill. TIME ALLOTMENT AND SPELLING EFFICIENCY.' ' ' It has been said repeatedly, on the basis of some investigation, that there is little correlation between the amount of time used for the teaching- of spelling, and the degree of efficiency attained in that subject. The following is an attempt to test that statement for the schools of Oakland, and to furnish the reader with the facts as they appear. One difficulty in passing final judgment, is that the degree of efficiency shown in any test is not the result of the work of one year, and since the time allotment varies widely between classes, often in the same school, we have no way of knowing whether the time which is being given to spelling by a given class this year is equal to, greater, or less than the amount used by that class the year preceding. It is true in ^^eneral, however, that, excepting grades 7 and 8, there is not a wide difference in the time allotment for the different grades. By using a large number of classes, therefore, we should be able to find the general tendency toward such correlation, if any exists. The actual amount of time set apart for the study and recitation of spelling in the different classes and grades is presented in table 18, which shows on the i-ight, the average number of minutes per week, and the per cent, of total school time devoted to this subject. From this table we see that classes in the different grades spend all the way from 15 to 125 minutes per week on the preparation of THE ADMINISTRATION AND TEACHING OF SPELLING 65 T 1 T V O -* T "f ^- » T 8 a-" T & ^ 6; [f s s £ «« ^^431^ 3: 5 >S ia ^ ^ NA C- SA 3 $ 1? c; ^ ^ ^ «> ■^ e U" -* ^ c- § »Z ^ ^s. N set 061 ^ •- sei m ^ '^ eJ 9L, •~. «o ^ -^ \n Oil — - S9I ■- - 091 <^ ■ — W 9SI «3 w "^i- QSI 5! 4- *• 5>- tc. * SU \ ■~ 0*/ -^ c^ - ~ In set '^ - «N •Si 00 aei -- «< to VBI ^ ■^ CM w> >^ >* (S~ vS > 3T cv< Vlll -- ■^ (Si "^ «^ 'i ^ VJ to f in a)' sn ■- "N to oil •^ -~ ■ — -. -^ c^ «^ «i «^i SOI -- - 001 w vS ^> ISo W CM <>< ■- «i ^*» io CO >■ ^ (x, ^ 5r Oo In 56 ■^ - - to (76 ^ '^ N ^>, - ^ to C^ CO C- fl ^ «> w S8 ri v- n -- ~- c* - - Cl 08 - "> W «0 ■*^ ^ ^ H ■~ >* N to 4- ^ cO to SL ^ c~ S9 ■^ ■X. - '^ lo -^ ~- ^ - 09 > ^-». ti t- 5*- «^ * — 3- Cj ss ^ ■^ -^ rt - - «i - e< t- « cj OS «>» * S5 CO 5 9: 0- r* o CO ^ ^ :5. ^ S^ 5? CO ^ - ^ '^ ''J -3 -■ -- to ^ CO -^ Si - '^^ :i- - '^^ -a co ^^ I5< 10 S£ ^ •>- — 10 ^ "> ■~ ro OD N N '!^ 0£ Tn «S 5^ - - - Q i^ -s i- -S - lo Ifi -- (:i to SI ^ w> ^ s^ t~ ^ ci «i 5^ 3- CO t>- s - ~ ot ~ ■^ - **^ ^ '-^ ^ t- «< ^ er S S{ c^ 0!. ^ S c<5 1 app-cg N ^ H fel N N i M w H N H H M ^ H M K h .ic •""'-L ^y^s i -a-^ij_ ii..4r4ya»^ ■""'X '•'•■•if^s i*jkI| 66 OAKLAND SPELLING INVESTIGATION spelling lessons, with an average in all grades of 60 minutes per week, which equals 4.69 per cent, of the total school time.' As be- tween grades, the range is from 43 to 67 minutes on the average. As to recitation time, the range between classes is from 10 to 110 minutes per week, with an average in all grades of 49 minutes, and a range between grade averages of from 38 to 55 minutes per week. Taking all the classes which reported both items, we find that from 25 to 200 minutes per week is devoted to spelling by different classes, that the average time is 103 minutes, or 8.53 per cent, of total school time, with a range between grades of from 82 to 118 minutes. If we are not surprised at this wide range in time allotments, we must at least be surprised to find that grade 2 actually spends a larger percentage of its time on spelling than is spent by any other grade, and that the 3rd grade comes next. The 2nd grade spends on an average more minutes per week in preparing spelling lessons than are spent by any other grade. It may be interesting to arrange the grades in order, first of the per cent, of time spent per week, and second in order of their rank in the test. Placing those which rank highest first, they are as follows : Order as to time spent, they are grades : 2, 3, 4, 5, 6, 8, 7 Order of standing in test : 5, 6, 7, 8, 4, 3, 2 If there is any correlation in this rough way, it appears to be of a negative sort. If this amount of time was given to spelling in the last six weeks of grade 1, then, when grade 2 took the test, the A division bad had approximately 18 weeks of work with an average of 101 minutes per week spent on the subject, while the B division had had one semester more than this. From which it appears that after all, the children are receiving quite substantial training in spelling be- fore they need it in their composition work, which in the third year is partly written. This appears to show that the standard set by this test, based on actual achievement in nearly 100 cities, is too high to be at- tained by a class which has had but 18 weeks' work in spelling, and that it can be attained by children who have had approximately one year of work (the upper half of grade 2, with 1 semester plus 18 weeks of work, made an average of 69 per cent., the standard being 70 per cent.). Certainly 8 :52 per cent, of a child's second year in school is enough, in fact more than enough, to spend on the single subject of spelling. If spelling does begin a little late in the Oakland schools, it begins very much in earnest when the time comes for it. Diagram 17 shows the points of emphasis in this distribution for study and recitation time separately and in combination. It THE ADMINISTRATION AND TEACHING OF SPELLING 67 Pi A e 1 o U H a o ■73 a a o o o pei 13 CO a o 68 OAKLAND SPELLING INVESTIGATION appears that 30, 50, and 75 minutes per week for recitation; 50, 60, and 75 minutes for study, and 100, 125, and 150 minutes per week for both are the most popular plans of time allotment. The distribution of time with respect to the number of periods per week given to the subject varies somewhat as appears in table TABLE 19. Showing the Distribution of Classes in Each Grade with Respect to the Number of Spelling Periods per Week, Periods per week GRADE 2 3 4 5 6 7 8 1 1 4 1 37 1 3 2 52 1 3 5 50 1 6 8 39 1 1 7 17 32 6 14 5 11 1 2 2 3 17 4 2 5 3 19, but there is a clear tendency to have spelling five times a week in grades 2, 3, 4, 5, and 6, with 3 periods per week as the central tendency for grades 7 and 8. TIME ALLOTMENT COMPARED WITH THAT IN USE IN OTHER CITIES. But two of the ten cities for which data was available devote a larger percentage of their total school time- to spelling than does Oakland, as is shown in table 30. The extreme range among TABLE 20. Comparative Standing of Ten Cities With Respect to the Per Cent, of Total School Time Devoted to Spelling.* City % of total time used for spelling Kansas City Cincinnati Oakland Milwaukee Salt Lake City.. St. Louis Cleveland Indianapolis Rochester San Francisco.... 10.70 9.57 8.52 8.07 ■7.11 6.39 5.94 5.33 5.38 5.23 * Data for all cities except Oakland and Salt Lake City is for the year 1907-8. THE ADMINISTRATION AND TEACHING OF SPELLING 69 classes in Salt Lake City is from 30 to 300 minutes per week, while for Oakland it is from 25 to 200 minutes per week. Some varia- bility between grades, and between schools is of course necessary, but it does not seem possible that 200 minutes per week, which is over 16 per cent, of the entire school time, could be used to advan- tage in the teaching of spelling. Such studies as have been made of this subject, make it doubtful whether more than 15 minutes per day should be used for spelling. For Oakland that would be slightly over 6 per cent, of her entire school time, which would place her about half way between the extremes represented by the ten cities cited above. A reduction of time so that it will not ex- ceed 100 minutes should certainly be recommended for the schools of Oakland. RELATION OF TIME CONSUMED TO ACHIEVEMENT IN THE TEST. Records for ten 8th grade, 15 6th grade and 14 3rd grade classes from 16 different schools were examined with a view to showing the relation between the amount of time the class is using for spelling and the average score attained by the class in this test. (B or A.B. classes were used because they have followed these time schedules for the past half year at least.) These records are shown in table 21 which presents the average scores for the classes, ar- ranged in descending order by grades, and the score for each class in the adjoining column. TABLE 21. Relation Between Minutes of Time Per Week Spent on Spelling and the Standing Made in the Test. 39 Classes from 16 Schools. GRADE VIII GRADE VII GRADE III Standing Time Used Standing Time Used Standing Time Used 86.3 90 90.3 125 88.8 70 812 40 90.0 100 85.1 120 80.0 100 88.4 80 82.0 125 79.8 70 87.6 60 78.4 95 79.3 60 86.7 165 73.7 120 78.2 180 82.9 135 73.7 120 75.4 60 81.3 115 73.7 80 74.9 90 79.5 150 71.0 80 72.0 90 79.0 150 70.3 125 68.9 90 77.3 100 68.0 120 77.1 100 65.4 30 77.1 150 60.0 150 73.3 95 59.1 125 71.7 90 57.7 80 71.5 100 In grade 8 the class using the least amount of time made next to the highest score. Similarly in the other two grades there is no apparent influence of large and small amounts of time. Diagram 70 OAKLAND SPELLING INVESTIGATION DIAGRAM 18. Showing Relation Between Class Standings and the Amount of Time Used for Spelling. From Records of 39 Classes, of Either B or AB Grades, in 16 Schools, Grades 8, 6, and 3. Ttrrie TT- -nr TJ- TT TJ- -JTT 18 presents the records of these 39 classes graphically,, from which it is evident that classes which use the most time do not receive correspondingly high scores. In fact if this curve were smoothed out, it would be approximately parallel with the base line, which means that there is no correlation either positive or negative be- tween the two sets of facts. Whether this proves anything or not it represents a cross section view of the situation as it is, and suggests that it would be worth while for some school — all schools for that matter — to establish a definite time schedule, say approximately 75 minutes per week, with less in grades 3 and 3, and mlaintain that schedule for a number of years, keeping careful record by standardized tests of the prog- ress made from grade to grade. There is no amount of theorizing, or single studies such as the above, which can ever take the place of such controlled experiment as is here suggested. Such a care- fully controlled study is a legitimate experiment to make in a school and the principal is the one with the best opportunities to make it and report the results. This does not argue that every school could wisely use the allotment best suited to one given school, but one such time schedule interpreted in terms of the results produced through a series of years would be a definite basis from which other schools and other cities might find appropriate standards. It is folly to use 200 minutes if 60 or 75 minutes will give the same results. THE ADMINISTRATION AND TEACHING OF SPELLING 71 With the development of scales for measuring- spelling efficiency we need to proceed to the task of standardizing the time allotment, the content, both as a whole and by grades, and the teaching of the subject. CONTENT OF THE COURSE IN SPELLING. In the information blank filled out by teachers, inquiry was made as to the sources from which spelling content is secured. From the replies to these questions it is clear that practically no teacher de- pends solely upon a spelling text. Perhaps in all a half dozen teachers do not go outside of a speller for word lists. In almost every case teachers are correlating the spelling with other subjects. The nature of this correlation is indicated by table 33, which shows the chief sources from which spelling lists are made up, and the number of teachers in each grade who draw upon these sources. TABLE 22. Showing the Chief Sources, Other Than Text, From Which Spelling Lists Are Selected, and the Number of Teachers by Grades Who Make Use of Each of These Sources. Source History Geography ... Reading and Language .... Grammar 23 20 26 9 26 32 30 12 Nature Study Arithmetic All written work 15 23 43 38 15 11 12 19 46 59 55 13 21 23 25 31 44 43 5 9 8 8 47 4 16 1 37 Other sources such as words used in conversations, words from street signs, names of things in common use in the kitchen, school- room, .etc., were mentioned, which means that various types of in- dividual and class lists are collected and made use of. This is an excellent practice, and could well be extended to the point of try- ing to work-out standard grade lists of words which are regularly found to be troublesome words. Judging from lists of words which teachers prepared for examination purposes in spelling, these sources were drawn upon rather liberally. There is danger in laying too much emphasis upon technical terms from other subjects. There were in the lists referred to here a number of geographical and physiological terms which children, or adults either, will rarely use. 72 OAKLAND SPELLING INVESTIGATION LENGTH OF LESSON ASSIGNMENT. In late years the tendency to devote less and less time to the formal study of spelling has been accompanied by a similar ten- dency in regard to the number of words assigned per lesson. A record of the number of words customarily assigned by 330 teachers of speUing in Oakland is presented by grades in table 33. TABLE 23. Showing by Grades the Variability in Length of Daily Lesson Assignment. No. of Words Assigned 2 Total 40 to 50.... 35 to 39.... 30 to 34.... 25 to 29.... 24.... 22..=. 20.... 18.... 17.... 16.... 15.... 14.... 13.... 12.... 11.... 10.... 9.... 8.... 7.... 6.... 5.... 3 to 4.... Average. 1 2 3 15.6 1 1 1 14.7 3 2 4 1 1 21 6 1 2 7 1 2 1 2 1 3 4 12 3 1 1 13 1 3 2 7 19.5 16.5 5 1 2 3 9 2 2 3 3 18 1 2 1 5 12.2 1 5 2 26 3 2 1 3 10 1 9.5 1 5 6 3 11 5l^ 5.5 3 3 4 10 1 8 57 12 7 8 40 3 6 19 14 65 5 14 9 7 29 6 13.3 The range in number is extremely wide, varying from 3 to 50, and running as high as 20 in grade 3. The average for the, differ- ent grades, as shown at the bottom of this table, varies, but it is safe to say that in the light of the best practice, all the averages are too high. There are two separate tasks for the teacher, one is teaching and the other is testing. These processes should not be. confused for neither will do the work of the other. An occasional test, covering a list of 20 or, rarely, 50 words may be desirable, but should be regarded only as a means of discovering whether or not the previous teaching of those words has been effective. From the length of lists indicated here but two possible conclusions can be THE ADMINISTRATION AND TEACHING OF SPELLING 73 drawn, either there is little teaching and much testing of spelling, or else the assignments include many words already familiar to the class. The latter is useless and time consuming, and the former is extremely bad pedagogically. This table does not include the number presented as review words. METHODS AND DEVICES IN USE IN THE TEACHING OF SPELLING. A part of the information which teachers were asked to give was a brief description of the rnethods and devices they are now using in the teaching of spelling. Many of the answers were very full and clear statements of procedure which is based on the best pedagogical theory. A few, however, were brief statements of the traditional ideas of bare assignment, followed by an oral or written test. To summarize these replies will of course necessitate much abbreviation, but it is thought that such a condensed statement of how a city teaches its children to spell may have some value. It would seem from a review of these replies, that every conceivable device is in use in some school. A few quotations from the replies of 2nd grade teachers will show effectively how spelling is being taught in that grade. From these we have the following, which are fully typical of the 37 replies from grade 3 : 1. "Spell word orally, trace it in the air, shut your eyes and see mental picture, then write it on the board." Pick out phonetic parts and silent letters, and spell words in parts, not .separate letters." 2. "Must know how to pronounce words, I teach spelling of familiar phonograms — then words containing these phonograms." 3. "All study and recitation is done with the words on the blackboard. After studying the words, pointing out the parts known (phonetic groups, consonants, etc.) and how they resemble or differ from words known, the children spell orally, or write indi- vidually the word. They are never allowed to guess, if not sure, look at the model, but all try to see who can be the first to spell correctly without referring to board. As these words occur in reading, writing or language they are also spelled." 4. "New words examined for parts known. Especial study of parts unknown. Writing words in air. On desk. On board. Finding words with known parts in reader." 5. "The words are presented, traced, then spelled. No hesita- tion allowed, to avoid guessing." 6. "Grouping the words containing similar elements. Oral spelHng by one pupil while class watches the list." One teacher of foreign children emphasizes "the use of words in 74 OAKLAND SPELLING INVESTIGATION sentences." Others try to get incorrectly spelled words out of sight as quickly as possible ; have but one word on the board at a time ; have all pronounce correctly as they look at the word; in testing, have all who can spell the words raise hands — others not permitted to write till more teaching is done on that word ; mark phonetic parts or silent letters with colored chalk; refer to the "u" in four, the "igh" in high, etc. as the keys which unlock these ■ words ; and systematic review of every word daily, weekly, and monthly, are other ideas brought out in the second grade replies. It is clearly evident that there is much real teaching of spelling in this grade, and that the traditional method of merely testing is the exception. In grade 3 practically all these ideas are expressed, and in addi- tion the following are typical : 1. "The words each child has missed the day before are given to him. He writes them on the board or on paper, or studies them silently. Each day I review the words most often misspelled." 2. "Read words in sentences — discuss meaning — written review every third day." 3. "I have had printed a list of somewhat more than 300 words and phrases in which they are required to be absolutely correct if these words are used in composition. List on good board in their seats." (Does not tell how these words are taught.) Other ideas, such as having the list of words for the lesson mem- orized, keeping a spelling record, contests, collecting individual and class lists, and copying misspelled words into individual note books are frequently mentioned. Again in the 4th grade practically all the above ideas enter with some reference to drill, and to the developments of rules for spelling- some words. In grade 5 some are making use of the dictionary, one says "cultivating the habit of noticing the spelling of new words encountered in reading," another : "I make sentences to emphasize difficulties as, he ate a piece of pie on the pier." Many have the children use the spelling words in sentences. In grades 6 to 8 more emphasis is placed upon definitions, analy- sis, drill, and some on diacritical marks. To show how generally these methods and devices are used the following rough tabulation is presented for grades 3 to 8 from which it is clear that nearly every teacher, of the 286 reporting, had ideas that were possible to classify under one or more of. the head- ings used. Naturally in the brief space, many did not mention.,every feature of their work. Certainly written spelling is more common in the upper than in the lower grades, but from this table it does not appear so. Such a common practice would likely be taken for granted. Aside from this the table is suggestive. Oral spelling THE ADMINISTRATION AND TEACHING OF SPELLING 76 TABLE 24. Distribution of Teachers With Respect to Methods and Devices Used in Teaching Spelling. Oral spelling Written spelling. Develop the meaning of word Trace the word in air or on desk Copy the list of words Study words by syllables and parts Have child use word in sentence Note unfamiliar and diffi- cult parts of word Emphasis on correct pro- nunciation 1 Development and applica- tion of rules Total No. of replies.... 35 32 15 18 14 13 17 36 35 16 37 5 7 16 17 18 25 42 20 2 9 21 23 27 38 51 63 10 37 20 1 3 16 13 17 19 1 54 7 20 17 2 13 14 13 38 4 37 5 11 17 2 7 20 8 18 5 32 Total 81 179 94 28 39 85 104 108 163 10 386 gradually decreases in amount from the lower to the upper gj-rades, tracing the word in air or on desk disappears by grade 6, detailed study of word by syllables decreases similarly, and use of rules re- ceives little attention even in upper grades. From these replies, really wooden teaching is conspicuously wanting though in cases bare study and drill are still believed in. These few, it must be said, are greatly in the minority, and they should realize that while the old idea of bare drill under pressure of "get these words or remain after school and write them a hundred times" will get results it will rarely get permanent results, and that such teaching is a lazy way of dodging responsibility which too often leads the child to dislike spelling, and school in general, in- stead of disliking his teacher with her mediaeval ec|uipment, as he really should do. In the replies there is little effort to distinguish methods front devices. It is true that the line between the two is hard to draw, for some devices may be applied in a variety of cases. The princi- ple underlying the nature and use of each device is clearly a detail in the statement of the relation of the child's mind to the subject matter in question, which is a statement of detail in what has come to be termed special method. Device is a term used to indicate the special plan or set of conditions under which, or objects by which, 76 OAKLAND SPELLING INVESTIGATION that principle is permitted to work. "Colored chalk" is a device when used in directing attention to a certain difficult part of a word, so also is the underlining of "ie," in chief, in order to focus the attention of the child upon the probable difficulty he will meet in trying to spell that word. Method is the process by which experi- ence (it may be spelling a given word) is accumulated. It is the way the mind or body acts in achieving an end, while devices are means or instruments whereby that process is facilitated. It may be that for the classroom teacher a technical use of these terms is unimportant, but if a knowledge of their meaning would obviate the too frequent assumption that a given device, a mere incident in instruction, is of fundamental importance in all classes and subjects, it would be worth while to understand them. There is of course much in these replies that does not come out in this brief space, most of which goes further to show that indi- viduality plays a large part in matters of detail, but that spelling is really taught and really studied intelligently. Most of the teach- ers who use the word "concentration" and "drill" make it clear that these are made effective by proper direction. CONCLUSIONS. In conclusion it must be said that the general policy for adminis- tering the subject of spelling is sound in its plan of relating spelling to other subjects. Similarly there can be no theoretical exception taken to the time for beginning this subject, which the schools are following, though the practical arguments that it should begin earlier are found in the poor results of the test in grades 2 and 3. To this objection it- must first be said; that it applies to the lower divisions of these grades only, and that from grade 3B on there are no evidences of either poor foundation work or low spelling efficiency. It is of course not evident whether these later achievements are or are not at the expense of other subjects. Opinion that they are would be offset by opinion that equal or greater gain is obtained in the earlier years when formal training is more difficult and trying for the child. The point is, there is little but opinion as yet upon which to base a decision as to the best time for beginning spelling, or any other subject. The tendency today is distinctly in the direction of defer- ring the formal work in number, writing and spelling, giving the first and in many of the better schools, a large share of the second year to reading, oral expression and dramatization. Psychology and physiology sanction, and tradition opposes this tendency. There is doubtless a best time to begin spelling, and our hope of finding that time lies in the growing tendency to give quantitative statement to the results of all teaching. Oakland's best proof that more good than harm comes from their present plan lies in showing superior THE ADMINISTRATION AND TEACHING OF SPELLING 77 achievements in other subjects which may offset the low grade of spelling work in the first three years. As to the amount of time given to spelling, the present practice needs revision in the direction of less rather than more. Spelling in the 7th and 8th grades should be carefully watched by teachers .and much incidental work on the subject should be done, such as having children make lists Of their own misspelled words wherever they are found in written work. But the amount of formal class attention to the subject should be small. Any evidence we have here, and that from other investigations, tends to show that much time is wasted on spelling. Fifteen minutes a day has come to be regarded as the maximum amount. Lesson assignments need to be materially shortened in all grades. At the rate of 2 new words a day from the 2nd grade through the 8th, the school would provide a child with something like 3800 words. If this number were gradually increased through the grades from 2 to 8 or 10 for the 8th grade, the child would leave school equipped with more words by far than a large percentage of children would ever have use for. In method there is much to commend, and the only suggestion is that present practice should be backed up by a wider reading knowledge than is evident in the replies to the question bearing upon the literature on the subject. APPENDIX A. A SUGGESTED READING LIST FOR TEACHERS AND PRINCIPALS. The following list of books and articles, including the best litera- ture on the subject of teaching spelling, should be familiar to every " teacher of the subject: 1. Suzzallo, Dr. Henry— The Teaching of Spelling, Houghton, Mifflin Co., 1913. 2. Cook & O'Shea— The Child & His Spelling, Bobbs-Merril Co., 1914. 3. Charters W. W. — Teaching the Common Branches, Chap. I, Houghton Mifflin Co., 1913. 4.. Jones, W. F. — Concrete Investigation of the Material of English Spelling. Univ. of South Dakota, 1913. 5. Boston, Dept. of Educ. Investigation, Bui. 1, Provisional Minimum & Supplementary Lists for Grades 1 to 8. Dept. of Educ. 6. Cook, N. A.— Shall We Teach Spelling by Rule? Jr. of Educ. Psych. 3 : 316-25. 7. Rice, J. M. — The Futility of the Spelling Grind, Forum, 1897. PP. 193-409. 8. Suzzallo & Pearson — Comparative Experimental Teaching of Spelling. Teachers' College 'Record, Jan., 1912. 9. Ayres, Leonard P. — The Spelling Vocabularies of Personal and Business Letters. Russel Sage Foundation, Pub. E. 126. 10. Bailey-Manly Spelling Books (2 vols.) with Teachers Man- ual, Boston, 1908. 11. Nicholson, Dr. Anne — "Speller for the Use of Teachers of Cal." State Printer, Sacramento, Cal. 12. Hicks-Champion Spelling Book, Am. Bk. Co., New York, 1909. In addition to this list, principals will find the following list in- valuable : 1. Pearson — Scientific Studv of the Teaching of Spelling, Jr. of Ed. Psych. 2: 241. 3. Rusk — Analysis of Spelling Errors of Adults. Jr. of Exp. Pedagogy, June, 1913. 3. WalHn, J. E. W.— Spelling Efficiency, Baltimore, 1911. SUGGESTED READING LIST. 79 4. "Has the Drill Become Obsolete ? Jr. of Ed., Psych. 1910 ; 200. 5. Abbott — Analysis of Memory Consciousness in Ortho- graphy. Psych. Rev. Monograph, 1909. 6. Burnham' — Hygiene and Psychology of Spelling. Ped. Sem. 13 : 471-503. 7. Buckingham — Spelling Ability, its Measurement & Distri- bution, Teachers College, Col. Univ., 1913. 8. Cornman — Spelling in the Elementary School, Boston, 1902. 9. Gregory, B. C. — The Rationale of SpelHng. El. Sch. Tr. 8: 40. 10. Kline — A Study in the Psychology of Spelling. Jr. of Educ. Psych. 3: 380. 11. Ayres, Dr. Leonard P. — ^A Scale for Measuring Efficiency in SpelHng. Russel Sage Foundation, New York, 1915. 12. Longnecker, Gertrude — The Teaching of Spelling, Bui. No. 3. Vol. n, San Diego State Normal, Cal. State Printer, 1914. Cornell University Library arW38761 Spelling efflclencey in the Oaldand sclio 3 1924 031 784 451 oiin,anx f^^^- ,^.^ -4 ^, "rn^-j- ^■y^'Vr^:' »,,^ K-» ^\ -jy >jm ^■■m^ 'W: