fd Public Health Service Publication No. 1000-Series 2-No. 24 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C., 20402 - Price 45 cents NATIONAL CENTER| Series 2 For HEALTH STATISTICS | Number 24 VITALand HEALTH STATISTICS DATA EVALUATION AND METHODS RESEARCH a study of the Achievement Test Used in the Health Examination Surveys of Persons Aged 6-17 Years A methodological study of the Wide Range Achievement Test, one of the measures used in the Health Examination Survey of children (Cycle 11) and of adolescents (Cycle II). Washington, D. C. June 1967 U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service John W. Gardner William H. Stewart Secretary Surgeon General NATIONAL CENTER FOR HEALTH STATISTICS FORREST E. LINDER, PH.D., Director THEODORE D. WOOLSEY, Deputy Director OSWALD K. SAGEN, PH.D., Assistant Director for State Relations WALT R. SIMMONS, M.A., Statistical Advisor PHILIP S. LAWRENCE Sc.D., Planning Officer ALICE M. WATERHOUSE, M.D., Medical Consultant JAMES E. KELLY, D.D.S., Dental Advisor LOUIS R. STOLCIS, M.A., Executive Officer DONALD GREEN, Information Officer PUBLIC HEALTH OFFICE OF HEALTH STATISTICS ANALYSIS [WAO M. MORIYAMA, PH D., Director DIVISION OF VITAL STATISTICS ROBERT D. GROVE, PH D., Director DIVISION OF HEALTH INTERVIEW STATISTICS Erijan L. WHITE, AM., Director DIVISION OF HEALTH RECORDS STATISTICS MONROE G. SIRKEN, PH.D.; Direct~r DIVISION OF HEALTH EXAMINATION STATISTICS ARTHUR J. McDOWELL, Directrr DIVISION OF HEALTH RESOURCES STATISTICS SIEGFRIED A. HOERMANN, Director DIVISION OF DATA PROCESSING LEONARD D. Mc GANN, Director Public Health Service Publication No. 1000-Series 2-iNo. 24 Library of Congress Catalog Card Number 67-60022 FOREWORD The psychological programs of the Children's Health Examination Survey (Cycle II) and the Ado- lescent's Health Examination Survey (Cycle III) aim at providing information concerning the number of psychological problems which exist in the Nation's noninstitutionalized population of persons aged 6 through 17. Achievement testing, therefore, was conducted not to evaluate achieve- ment per se, but because many developmental and psychological problems first come to the attention of teachers, psychologists, physicians, or other caretakers as "achievement problems." Because of the survey nature of the operation, no one health factor, whether physical, physio- logical, dental, or psychological, can be evaluated as thoroughly as it would be in a nonsurvey setting. As a result, most of the measurements are collected using either specially designed tech- niques or abbreviated forms of widely used, longer procedures. By means of methodological studies these special or abbreviated instruments are then evaluated to see what relationship exists between them and established, criterion measures. The instrument chosen for measuring achievement in reading and arithmetic in the Children's and the Adolescent's Health Examina- tion Surveys was the 1963 revision of the Wide Range Achievement Test (WRAT) originally pub- lished by Joseph Jastak in collaboration with Sidney Bijou in 1946, The WRAT is notan abbrevi- ated version of a longer, well established test, nor was it specially developed for use inthe survey. It is a hitherto relatively unproven short test for the rapid assessment of achievement skills. It was KR409 Us12 no.24-2"] PUBLIC HEALTH LIBRARY selected because of its brevity and also because it was held by many clinicians to be a good predictor of performance on the more traditional achievement tests. Because of the nature of the WRAT, a study was designed to establish the relationship between it and the Stanford Achievement Tests for individ- uals in grades 1 through 9 and the Metropolitan Achievement Tests for individuals in grades 10 through 12. Hopefully, a description of this re- lationship will permit the reader to evaluate our forthcoming reports dealing with the incidence of underachievement in the Nation's population of persons aged 6 through 17. In addition, scientists will have available, for the first time, information concerning the re- lationship between the Wide Range Achievement Test (reading and arithmetic sections) and ap- propriate subtests of the Metropolitan and Stan- ford Achievement Tests. For a test originally published in 1946, such a study is long overdue. This study is the product of contract number PH 86-65-52 between West Virginia University and the National Center for Health Statistics. The project director was K. Warner Schaie, Ph.D., professor of psychology, West Virginia Uni- versity; and 1 was the project officer. Contri- butions by the examiners and other project personnel are gratefully acknowledged. Their names are listed in Appendix III, Lois R. Chatham, Ph.D. Psychological Advisor Division of Health Examination Statistics 704 CONTENTS Page Foreword === moomoo mmm i Introduction == === mmm meee meee 1 I. Design of the Study--=------=-ccmcmo mma 2 General Format----=--=- mmm mmo meee 2 Subjects =-----mmmmm eee mmm 2 Monongalia County Sample------- mmm em 2 Control Sample --- c= c-mmmm eee e 2 Socioeconomic CharacteristiCS---=-=-=c-mmmmmmccmme cee ceeme eo 4 General Ability Level---mcmemmm oe eee eee 6 Test INStrumentS------ mmc mo ee eee 8 Wide Range Achievement TeSt-----==--m-commmcmmcm emcee meee 8 Stanford Achievement Test--==--mm o-oo eee 9 Metropolitan Achievement TeSt--=--=-coom ome emma 10 Examination Procedure-----=c- mmm eee oo 11 Reliability of Scoring Procedure-----c comm mm meme 12 II. The Elementary School Study---=-===--=- comm mmo ema 14 Background Data=-----=c om mm mmm 14 Adequacy of Grade Level Placement------==--cmmcmmmcmcceceeme en 14 Performance on the SAT ---m cmc ee eee 18 Relation of the WRAT to the SAT----cmmm emcee 19 Discrepancies Between WRAT and SAT Grade Level Estimates------ 23 The WRAT SectionS=-=-=- comme eee eee 24 Interrelationship--====- meme cme eee 24 Relation to General Ability and Socioeconomic Status------------- 26 CONTENTS—Continued Page III. The Junior High School Study----=---=--m-mm-mmoaeccmcmccmm cen om me 27 Background Data-----=----=--ccmoommo moomoo oomommoemom moo 27 Adequacy of Grade Level Placement--------====e-mmmoomoooooooo- 28 Performance on the SAT----eemmcmcmcccm mcm meme mcm mcmmme mmo 29 Relation of the WRAT to the SAT-=--cecmccccmccmm cece mmm 29 Discrepancies Between WRAT and SAT Grade Level Estimates------ 30 The WRAT SectionS----=-ceemmmcm ccc mmcmc ecm mec cme mmm mmo 34 Interrelationship =--=-==cccemeommoccmccme meme mm mm ememme eee 34 Relation to General Ability and Socioeconomic Status----==--=----~ 35 IV. The Senior High School Study-------=-e---mecemcmcccomcmeenmmne nme 36 Background Data------=---cemcemoommmmmmmmmeooooooeooome os 36 Adequacy of Grade Level Placement--------c--emc-mmmmmmmanaconm- 37 Performance on the MAT -cccmcmmmmmcmmcc cece ccm meme m meme 38 Relation of the WRAT to the MAT ccc rcmmmmmm mmc meme emcee 40 Discrepancies Between WRAT and MAT Estimates------------==--- 42 The WRAT Sections-------ccmcmecmmmmcmcmm ccc mmm mmm mmm momo 43 Interrelationship ---==cem-cmomcem momen mec em oe 43 Relation to General Ability and Socioeconomic Statug----==---=--- 43 V. Validity of the WRAT at Extreme Ability Levels----=--o-co-ooocommnn-- 43 WRAT Performance at Extreme Levels---ce-ecmeemccccmccececnn- 43 Relation Between the WRAT and the Criterion Variables-«--=------- 49 Discrepancies Between WRAT and SAT Grade Level Scores--------- S1 VI. Conclusions =-=-----cecemmmmmcmc mmm cece eens mm mmmmeee momo 51 Appendix I. Level I of the Wide Range Achievement Test Form Used in the Elementary School Study---===c-ceccmcomcmmommmmm mmm mmce oc mmm momen mes 53 Personal Data------ccececcercmmcmmm meme cee meee een me——eo oo 53 Test Instructions, Level I Reading---==--=-co-cceocmmmmmmmmmmaoacnn- 53 Pre-Reading Section----------cocmememmmme oom cm om mmm momen 54 Level I—Pronunciation Guide for Reading Test--------eemcccc-cnmeenmm- 54 Level I— Arithmetic =-e-emm--ecocccecmem mmc m ecm cee meme 55 CONTENTS —Continued Page Appendix II. Level II of the Wide Range Achievement Test Form Used in the Junior and Senior High School Studies - =~ mec c comme 56 Personal Data----- ccm eo. 56 Reading Instructions=--=-- = ceo... 56 Level I1— Pronunciation Guide for Reading Test--==ccmcceomcmcccmmmnan. 57 ArithmetiCed.evel Hl--mmmremcceme rca v emcee meee e———————— 58 Oral ArTHIMEUC—Level Huerees seven smminumne ness cmene nme ewe sans 58 Appendix III, Personnel Participating in the Data Collection and Analysis--- 59 IN THIS REPORT the suitability of the Wiae Range Achievement Test as a valida measure of school achievement for use on a national health survey is discussed. It was founa that the Arithmetic ana Reaaing sections of the 1963 Re- vised Wide Range Achievement Test have reasonably good construct validity as judged by their relation to the Stanfora ana the Metropolitan Achievement Tests. The WRAT was founda to be suitable for use with chilaren of widely differing socioeconomic backgrounds and different ability levels. The Arithmetic section was found to be valia at both high and low ability levels. The Reading section, however, was not suitable for high school students at the low end of the ability continuum. The valiaity of the WRAT as an estimate of grade level placement showed considerable variation. Level I of the Reading and Arithmetic sections has a tendency to overestimate actual grade level and achieve- ment as measured by the Stanfora Achievement Test. Level 1I of the Avithmetic section underestimates actual grade level but is a satisfac- tory estimate of criterion achievement measures. Level 1I of the Reaa- ing section tends to overestimate actual grade placement anda to undaer- estimate performance on the Stanfora Achievement Test for junior high school students. For senior high school students it tends to ovevresti- mate performance on the Metropolitan Achievement Test ana lo unaev- estimate grade level placement. In spite of the fact that the validity coefficients vary considerably, de- pending on the grade level and geographical region involved, there is sufficient evidence of substantial correlation with criterion measures al every age level investigated to consider the WRAT a satisfactory brief estimate of school achievement. SYMBOLS Data not available----=ccmcmcmmmmccenenam — Category not applicable------=-=-ecmcmoooo- Quantity Ze€ro-----=-=-=====ccc--ocom--mn- “ Quantity more than O but less than 0.05---- 0.0 Figure does not meet standards of reliability or precision--------=----==--- * A STUDY OF THE ACHIEVEMENT TEST USED IN THE HEALTH EXAMINATION SURVEYS OF PERSONS AGED 6-17 YEARS K. Warner Schaie, Ph.D., West Virginia University INTRODUCTION The purpose of this study is to assess the validity of the Wide Range Achievement Test (WRAT) in terms of its ability to predict grade level placement on the Metropolitan and the Stanford Achievement Tests (MAT and SAT), which are the criterion measures. Attention is given to the discrepancies which exist between the WRAT grade level ratings and performance on the criterion measures, in terms of grade levels. This was done by analyzing the rela- tionships which exist between the WRAT and the criterion measures. To control for the bias which might be in- troduced by the geographic location of a sample, one sample was chosen which consisted of a population of children in grades 1 through 12, all of whom were students in a single school system. Data from this sample were then com- pared with data obtained from a sample con- sisting of students from widely separated sec- tions of the country. Because of the nature of the population investigated, this study had been divided into three parts. Thus, after the general design, criterion measures, and selection of subjects are described, the results will be reported in detail, grouped separately for the analysis of the relation between the WRAT and criterion measures (1) in elementary grades, (2) for the junior high school population, and (3) for the senior high school group. In each instance, data and appropriate comparisons will be pre- sented based on children in the geographically homogeneous sample (Monongalia County) and on children in control samples from widely separated geographic regions. A technical study of the type here reported requires samples which should be reasonably representative of the general population. This does not imply that concerted attempts should be made to attain the exact replication of the population census or to provide random samples of the total population. It is of greater importance to ensure the adequate representation of groups at all levels of ability in order to be able to assess properly the success of achievement tests in evaluating typical as well as atypical performance. Considerable effort was directed, therefore, toward the objective of achieving representativeness by appropriate selection of samples. |. DESIGN OF THE STUDY GENERAL FORMAT The Arithmetic and Reading sections of the Wide Range Achievement Test (WRAT) formed the basic research instrument and were given to all subjects. A group-administered achievement battery was also given to each subject. The two group tests chosen as the criterion measures with which the WRAT was compared were the Metropolitan Achievement Test for use with grades 10 through 12 and the Stanford Achieve- ment Test for use with grades 1 through 9. With the Stanford Achievement Test the form given varied with grade placement. In addition to the Arithmetic and Reading sections of the Wide Range Achievement Test and the group achievement tests, information was collected on the socioeconomic characteristics of the pupils, and scores on general ability tests were recorded. SUBJECTS Monongalia County Sample The first sample selected was a relatively homogeneous school system chosen for the pur- pose of providing data concerning the efficacy of the WRAT across the different grades. The schools were selected to include the broadest representation possible of urban and rural chil- dren with a wide range of socioeconomic back- grounds. Schools having a marked concentration of university faculty children were not included in the sample. To achieve adequate representation and to permit separate analyses at each grade level, approximately 50 boys and 50 girls were selected from each grade level. Data for the Monongalia County, W. Va., sample were obtained in three elementary schools, each of whichcoveredgrades 1-6; one junior high school (grades 7-9); one junior-senior high school (grades 7-12); and one senior high school (grades 10-12), For administrative reasons, as well as to avoid the possibility that selection schemes might artificially truncate the distribution of talent in the sample, all children in the elementary schools, the junior-senior high school, and the junior high school were tested. Since the high school sample was predominantly rural, it was decided to supplement it by randomly selected cases from the University High School, which served an urban area. Here names were picked at random from the grade rosters until each grade quota was completed. Approximately 10 percent oversampling was conducted to provide some insurance against the contingency that some children were likely to drop out or fail to be available for either the individual test or the group test. The practical necessity of including entire classrooms in the testing procedures in some instances required the testing of some additional children. Tables 1 and 2 give the total number of children in- cluded in the Monongalia County elementary and secondary samples to whom either a group or an individual test was given as well as the number of children included in the final sample. These latter figures indicate the number of subjects on whom scorable records were obtained in both individual and group testing situations and on whom data are included in the statistical analyses. Control Sample In order to avoid the possibility of obtaining data which would reflect the peculiar circum- stances of a single homogeneous school system, additional data were collected on children in widely dispersed portions of the United States. Rather than testing smaller samples, it was decided to replicate the sample size but tocollect data on only four grades in eachof three different locations. Since the principal sample was collected in the mideastern part of the country, the control samples were placed in the midwestern, Rocky Mountain, and west coast areas. Control sample A covered the first, fourth, seventh, and tenth grades and involved two elementary schools, a junior high school, and a senior high school in Milwaukee County, Wis. The schools were selected so as to be at the Table 1. Number of elementary school subjects included in the Monongalia County sample and number on whom complete records were obtained, by sex and grade Grade Total | Boys | Girls | Total | Boys | Girls 0 Number with Dumber in sauple complete records Total elementary school sample------- 736 368 368 683 342 341 Grade l-==----mcmmmmmmmmm meme mm mm 116 56 60 114 54 60 Grade 2--=-==m--omommmmeeeeeoo ol J——— 117 62 55 111 59 52 Grade 3--==----m-mecmmmee eee Fem 121 60 61 113 54 59 Grade 4=-=-=m-c mmm eee 127 73 54 121 71 50 Grade S--=--=-m mmm meee eee 111 52 59 105 50 55 Grade 6--==-c=cmmmmmm eee 144 65 79 119 54 65 Table 2. Number of secondary school subjects included in the Monongalia County sample and number on whom complete records were obtained, by sex and age Grade Total | Boys | Girls | Total | Boys |Girls Number with Number in sample complete records Total secondary school sample-------- 706 355 351 633 314 319 Total junior high-=---c-cceeomomaaa oo 376 192 184 330 166 164 Grade 7-===-=- momma 125 12 53 111 61 50 Grade 8-=----m=-o ome 117 63 54 101 51 50 Grade 9----mecmmo mmm 134 57 77 118 54 64 Total senior high-=---=-cceocommmmmaa oo 330 163 167 303 148 155 Grade 10-===--ccc momma 109 54 55 97 48 49 Grade ll-=--c-omomm ao 110 56 54 103 51 52 Grade 12-==-ccmc mmm. LL) 53 58 103 49 S54 Table 3. Number of elementary school subjects in the control sample and number on whom complete records were obtained, by sex, grade, and location of sample Grade and location Total || Boys | Girls | Total | Boys | Girls Total elementary school sample------- Grade 1 (Wisconsin)-====--mecmooomoooooo oo Grade 2 (California)---===ccemooocmaoooo = Grade 3 (Colorado) --=====mmeocoommmcmcaaa = Grade 4 (Wisconsin)-===-=--meocmmmmcoaoaoon Grade 5 (California)---==--cemmoocmooooaoooo Grade 6 (Colorado) ---======-cmmecoooooooooo Number in sample Number with complete records 680 338 342 627 317 310 103 50 53 103 50 53 120 62 58 104 54 “50 113 60 53 104 56 48 104 50 54 100 50 50 123 55 68 110 50 60 117 61 56 106 57 49 3 periphery of the metropolitan area and thus are assumed to be reasonably comparable in socio- economic distribution to the other samples. Control sample B included the second, fifth, eighth, and eleventh grades and was collected in Duarte, a suburban semirural school district in Los Angeles County, Calif. This district also had some similarities with the main sample in that it had a small sprinkling of rural and minor- ity group children. Here, also, data were collected in two elementary schools, one junior high, and one senior high school. Control sample C, finally, covered the third, sixth, ninth, and twelfth grades and was collected in Fort Collins, Colo. Fort Collins is a college town close to rural and mining areas with a metropolitan area similar in size to the Monon- galia County situation. Again two elementary schools, a junior high, and a senior high school furnished the subjects for this sample. Tables 3 and 4 give the number of elementary and secondary school children in the control sample, and table 5 gives, by geographic location, the number of children who were included in the sample and for whom complete records are available. Socioeconomic Characteristics Parents’ occupations and students’ ability levels were determined in order to ascertain whether the sample selected actually covered a representative range and to permit appropriate statistical adjustment if necessary. Occupational level for the head of household was coded accord- ing to the following scheme: O - unskilled laborers! 1 - domestic laborers (including gardeners and janitors) 2 - operators (factory, and similar work requiring no special training) 3 - service occupations (including mailmen, service station employees, dry cleaners, etc., all requiring only limited training)! ISpecial cases—disabled and unemployed workers were classified as O, retired workers as 3, undergraduate students as 7,and graduate students as level 8. 4 - protective occupations (policemen, fire- men, guards, soldiers; however, ser- geants were classified as 6 and com- missioned officers as 8) S - craftsmen (including all trades requiring an apprenticeship or formal training) 6 - clerical and sales (excluding news ven- dors, grocery checkers, dime store clerks, who were classified as 3) 7 - managerial and proprietors (including independent farmer-operators; tenant farmers and farm laborers, however, come under classifications 1 and 0, re- spectively)? 8 - semiprofessional (including most occu- pations requiring college training but not more than 2 years of graduate work)? 9 - professional (all occupations requiring 2 or more years of graduate work, including lawyers, social workers, all college instructors, and school adminis- trators. Teachers and nurses would ordi- narily be classified in level 8 unless they have administrative positions) The scheme used is a modification of the major headings used in the 1950 census. It was first used in Measuring Behavioral Rigidity: A Fac- torial Investigation of Some Tests of Rigid Be- havior (K. Warner Schaie, unpublished M.A. thesis, University of Washington, 1953). The distribution of parents' occupations for the subjects included in the Monongalia County elementary school sample is given in table 6. It may be seen that the distribution was quite uniform throughout the six grades included in this sample and would seem to be reasonably representative of the socioeconomic structure of the local community. Table 7 gives a similar distribution for the Monongalia County secondary school sample. The distribution again was quite uniform throughout the six grades examined. There was, however, some underrepresentation at the upper level due to the fact that the area where most university people live was avoided. Tables 8 and 9 give the socioeconomic charac- teristics of the control samples. The distribution for the elementary school samples was similar to that obtained in Monongalia County. The Table 4. Number of secondary school subjects in the control sample and number on whom complete records were obtained, by sex, grade, and location of sample Grade and location Total || Boys | Girls | Total || Boys | Girls Number with Nusber in sample complete records Total secondary school sample-------- 791 402 389 596 291 305 Total junior high -==---cccceceaannao- 511 272 239 327 165 162 Grade 7 (Wisconsin)--=--=---eeccmmmcccceeeoo 104 51 53 104 51 53 Grade 8 (California)------c--ccecmcemca oo 128 65 63 109 57 52 Grade 9 (Colorado) -=--=-----cccmmmmmmmcaaao 279 156 123 114 57 57 Total senior high-===--cccccecaaaaao-- 280 130 150 269 126 143 Grade 10 (Wisconsin)-=--=---=ccmcommcanaoao- 103 52 51 102 52 50 Grade 11 (California)------cecmemmmeccaoaa- 106 52 54 98 48 50 Grade 12 (Colorado) ==-=----=-=---cccmcmamaaaao 71 26 45 69 26 43 Table 5. Number of subjects in the control samples and number on whom complete records were obtained, by sex and location of sample Location Total || Boys | Girls | Total || Boys | Girls Number in sample Number with complete records Combined samples--------=cccmcecaoao- 2,913 || 1,463 | 1,450 | 2,539|| 1,264 1,275 Total control sample-=---=---cceccecoeao- 1,471 740 731} 1,223 608 615 California-------c-cmmmcmm remem eee e em 477 234 243 421 209 212 Colorado===-=-=mmc mmm m meee = 580 303 277 393 196 197 Wisconsin-==--cemccmcc mmm meee 414 203 211 409 203 206 Total Monongalia County sample--------- 1,442 723 219] .1,3l6 656 660 Table 6. Number of elementary school subjects in the Monongalia grade of subject and occupational level of parent County samples, by Grade of subject Occupational level of parent Total|| 1 (2 [3 |4 |5 |6 Number of subjects 0-Unskilled laborers=-===-=a-cccomm mmo meeeee, 61 18| 9] 11 | 16 2 5 1-Domestic laborers--------ccceoommm eee 24 31 5 4] 2| 8 2 2-0peratives-==-== mmm meee eee 29 41 21 4) -| 9] 10 3-Service occupationg==--==--ceommo eee 119 18 1241723] 20 17 4-Protective occupations=-=-===-c-cooommmmm eon 22 41 51 1 7] 1 4 S-Craftsmen---------m omc o 146 || 14 | 27 | 2329] 28 | 25 6-Clerical and sales==--=====ccocmomcmm eo 80 15 918 | 11] 12 15 7-Managerial and proprietors=--------c--cooeomcooaao- 100 18 | 1517 [15] 12 | 23 8-Semiprofessional-==--====-mmccmmom een 63| 12 812 |14| 7| 10 9-Professional-----cmmccmmo meen 39 8 71 6] 4] 6 8 Table 7. Number of secondary school subjects in the Monongalia County samples, by grade of subject and occupational level of parent Occupational level of parent Grade of subject Total 7 8 9 10 [11 | 12 0-Unskilled iaborers------=---cc-=-cccoeuo-o- 1-Domestic laborers---------=-ec--ccccccco--- 2-Operatives~~==rrrewme-weemme mmm ccn nnn — 3-Service occupationg======wmmsansmmes ones 4-Protective occupations-----e-mweeemeenen=- 5-Craftsmen-------==---mec-ooeeeec cam —————— 6-Clerical and sales=»==m=remm-meeemen enon 7-Managerial and proprietors---------------- 8-Semiprofessional--=-----====cc-ccce--nooo-- 9-Professional-----=-=rmecmcmeece remem em Number of subjects mmm 118 211192517 |17 19 mm——————— 45 41 9 5] 5113 9 —mmmmmme- 19 41 31 9 2] = 1 mm——————— 119] 28 | 172122 |13 18 ed wesw 5 2 1] - 1 1 - ie mi 202 || 24263936 |43| 34 ———— 34 5 5| 6 31 5 10 ————————— 53|] 14) 410 8 7 10 a 2) 31 6 31 3] 4 2 ——mmmmma- 7 1] - “ & = 6 secondary school samples in the control group, however, tended to have higher socioeconomic levels, possibly suggesting different patterns of high school attrition. This was particularly noteworthy for the Colorado samples. The rural portions of the Colorado samples were likely to be children of farm laborers for whom high school dropout would be higher than for the West Virginia children. This factor resulted in a higher average socioeconomic level for the children who remained in the Colorado samples. General Ability Level The distribution of general ability in the sam- ples was studied by determining the score on the most recent group intelligence test which a given child had taken. This meant that scores were used on tests which had been given anywhere from 3 months to 2 years prior to the present study and that several different tests or test forms might have been utilized. Most scores, however, were from the California Mental Maturity Test Table 8. Number of elementary school subjects in the control samples, by grade of sub- ject and occupational level of parent Occupational level of parent Grade of subject Total|| L |2 [3 [4 |5 |6 0-Unskilled laborerS-------=-====--ecaccao--- 1-Domestic laborers-------c-cececccccoaaan=- 2-Operatives-------=-=-==ccscecmemacaeoanaoa- 3-Service occupations-------=-=cc-ccaocaanan 4-Protective occupations-------===c--ec----- 5-Craftsmen-=-=----===-e--mee-ccceco-cooo-on= 6-Clerical and saleS-=====ceceeacocccacc-co- 7-Managerial and proprietors---------------- 8-Semiprofessional------=c--ccococccacaonnn- 9-Professional--=c-cceccccmcamammacanoaaaan- Number of subjects meme 110 || 23 | 12222711} 15 me ton 7 11 311 2 - ————————— 19 51 1 -| 4) 7 2 mmemmm—— 114 91331201026 | 16 mmm 15 1] 31 2} 3] 3 3 cemm———— 139 || 30 {28 | 17 |21 22 | 21 mmm—————= 57 8 8| 8 |13]|1L 9 meee 73 9 (102110 9 | L&4 emma 700 17) 6} 910 |L17 | 12 -1 4 2) 2] 14 Table 9. Number of secondary school subjects in the control samples, by grade of sub- ject and occupational level of parent Grade of subject Occupational level of parent Total 718 9 10 | 11 | 12 Number of subjects 0-Unskilled laborers--------emom common eeeeooo- 60 8| 10] 11 [13 | 14 4 1-Domestic laborers------oee ooo 6 yp Lf - 1 2 1 2-0peratives=-= === mo moo ooo ee 11 1] 4] 1] 1] 4 - 3-Service occupations---------commmmm ee. 127 12] 25 34 | 15] 28 13 4-Protective occupations-=----mmecmomm mom mooo 11 1 5 2 1 1 1 S5-Craftsmen-----=cmc ooo meme aol 118 221 28] 14 [19] 20 15 6-Clerical and sales----===commcmmmm meee eo = 74 20 12| 1311] 10 8 7-Managerial and proprietors---------ecocoommmmao_ 106 25| 14| 18 | 23 | 16 10 8-Semiprofessional--=-=-ccocm momma - 52 11 9 914] 2 7 9-Professional-====-cecc momma 29 3 1) 11 | 4] - 10 and the Otis Group Intelligence Tests. Because intelligence tests. The meaning of the ability of the variety of intelligence tests which were levels used was as follows: used, it was decided that only gross classifica- 1 - mentally defective (IQ of 70 or below) tions were in order. Ability levels were therefore 2 - borderline (IQ of 71 to 80) recorded on a 7-point scale. Assuming that the 3 - dull normal (IQ of 81 to 90) tests used all had a standard deviation of 15 4 - average (IQ of 91 to 110) points, the intervals for the 7-point scale were 5 - bright normal (IQ of 111 to 120) set at intervals comparable to the descriptions 6 - superior (IQ of 121 to 130) being used for the interpretation of individual 7 - very superior (IQ of 131 and above) Table 10. Number of subjects inthe Monongalia County sample,by ability level and grade Ability level (IQ) Grade 9 1 71-|81-| 91- | 111- | 121- 151 below 80 | 90 110 | 120 130 above Number of subjects Totale-mecc emcee ieee eee em 21 | 43| 125] 577 236 63 15 Grade 3-=-=meccoc meee am 2 3 12 57 30 7 1 Grade 4-===cccccc mcmama 1 2 14] 65 24 9 6 Grade S====-cecm eee mao 3 5 6 65 19 7 - Grade b======----cmmc meee an - 3 9 59 35 10 3 Grade 7-==--=e-ccmccem mcmama eae 3 8 19| 60 14 7 Grade 8-===---cmcem meee ae 1 8 15| 54 20 3 = Grade 9-=--=--cccc cma em 2 4 | 16 74 18 3 1 Grade 10--=-mcemccmc cece cece 2 4 12 47 25 4 3 Grade ll-=+---ccccccaccce ccc cece rec cce emma 4 4 11 48 27 9 = Grade 12-----ee-mmcmceeemem ammo aeeae 3 2| 11] 48] 24 4 1 Table 11. Number of subjects in the control samples, by ability level, location of sample, and grade Ability level (IQ) Location and grade 70 171] 81- | 91- | 11-121] 2H selow 80 [90 | 110] 120 | 130 | ,pove Number of subjects Total----=--c--ommmmc mmm mmm mmo 1| 25 71 | 582 | 318 | 172 Lb California----------cccmmmcm meme m meee = - 2 48 | 245 75 26 8 Colorado---=---===c-cecmmme mmm meme me = 1 19 8 | 164 103 79 24 Wisconsin---=----=---mememmmce meme - 4 15 | 173 | 140 67 12 Grade l--------c-cmmmem mmm mmm mmm m m= - - A 44 41 16 1 Grade 2-------m-m--mmmemmmeme meme me momo - 3 6 | 65 20 7 3 Grade 3--------c-mmmmmmm meme mmm - 1 1] 62 31 9 - Grade 4------=----mommmmm meme meme m mm mem - 1 6 38 30 15 10 Grade 5----------emmmme meee meme me - 5 13 | 56 24 9 3 Grade 6----=--==-=-=--emememmemememe moomoo 1 2 31 35 23 32 10 Grade 7--------m---mmmmme mmm mmm meme mem - 1 3| 44 33 22 1 Grade 8------------mmmmmmmmme meme mm - 5 12] 67 18 5 2 Grade 9----------------meemmm—eemcmmmem mmo - 1 3 | 44 33 23 10 Grade 10----------memmm mmm mmm meme meee - - 5| 47 36 14 - Grade 1ll--------=------mmmccccee emma —— - 6 17 | 57 13 5 - Grade 12--------------mmmemmmmmmmmmmme mma = - 1] 23 16 15 4 The distribution of general intellectual ability for the Monongalia County samples is reported in table 10 separately for each grade and for all grades combined. However, no ability scores were available for the Morgantown samples in grades I and 2. Similar data for the control samples are given in table 11. A basically symmetric distri- bution extending to both extremes was obtained for the Monongalia County samples, although there was some upward skewing due to greater inclusion of children from higher socioeconomic levels than had originally been anticipated. This skewing was even more pronounced for the control samples in whose school districts policy decision leads to assignment of children of low ability to special classes both earlier and more systematically than is the case in Monongalia County. The skewing was most pronounced for the Colorado samples and least pronounced in the Wisconsin samples. The distribution of children in the California samples was fairly similar to that found in Monongalia County. Patterns across grades were fairly uniform within each geographical area. While these samples are certainly not exact replication of the distribution of talent within the population, they would seem to be broadly representative of typical school populations; thus they meet the sampling requirements set forth in the introduction to this report. TEST INSTRUMENTS Wide Range Achievement Test The principal instrument used for this study was, of course, the 1963 revision of the Wide Range Achievement Test, for which validity data were to be obtained. Because of the purpose of this study, the parts of the WRAT administered were confined to the ones included in the Health Ex- amination Survey, i.e., the Reading and Arith- metic sections. Two levels are available for each of these sections in the 1963 revision of the WRAT. One is designed for primary school chil- dren and the other for secondary school children. In this study one or the other form was used, depending on the appropriate grade level. The Arithmetic section of the Wide Range Achievement Test consists of a series of written arithmetic problems ranging from simple addition and subtraction through algebraic problems. Al- though defined as a timed test, it is a power test in the sense that the outside time limit of 10 minutes amply permits the students to work up to the maximum level of their arithmetic skills. The Reaaing section of the test consists of a list of words ranking from very simple ones such as "cat," "go," and "in'' to complicated ones such as 'belligerent" and "occurrence." It is assumed that the student who fails to recognize a given word is likely to mispronounce it also. The test, nevertheless, is not one of pronunciation or diction, and speech defects or colloquial usages are not penalized. For students at the very low level of ability the Arithmetic section contains an oral part and the Reading section contains a preword part involving letter recognition. The criterion measures used were the group achievement tests. These tests were the Stanford Achievement Tests in the appropriate form, de- pending on the grade level, for grades 1 through 9 and the Metropolitan Achievement Test for grades 10 through 12. Each of these group achievement test batteries contains subtests whichare directly pertinent as validating criteria for the WRAT. In addition, they contain other subtests covering school performance, whichis less directly related to reading or arithmetic. In designing this study it was required that certain tests of immediate relevance as criterion variables be routinely administered, while the other subtests could be administered at the discretion of the partici- pating schools. As a result the minimal amount of required data is reported at all grade levels while additional, or complete, data on the criterion batteries vary from one grade to the next, de- pending upon the discretion of the schools. Stanford Achievement Test The specific forms of the Stanford Achieve- ment Test (SAT) which were used are as follows: Primary 1, Form W, for grade 1; Primary II, Form W, for grades 2 and 3; Intermediate I, Form W, for grade 4; Intermediate, Form J, for grades 5 and 6; and Advanced, Form Km, for grades 7, 8, and 9. Not all SAT forms have the same number of subtests. Thus, six subjects are covered at the first grade level, eight at the second and the third grade levels, ten at the fourth grade level, and nine at the fifth to ninth grade levels. The tabulations for data relating to SAT have been arranged to give maximum com- parability from one grade level to the next. Missing data indicate subjects for which no SAT subtest was available at a given grade level because the particular school did not elect to administer the optional tests. The following paragraphs describe the subtests of the criterion batteries and their contents. Two SAT subtests are airectly relevant cri- terion variables for the Reading part of the WRAT: Word Meaning or Vocabulary (graaes 1-9). — The Word Meaning, or Vocabulary, test employs a multiple choice type of item in which the pupil is required to select the proper answer for a given stimulus word from a series of three or four alternatives. This is essentially a word recognition test, Paragraph Meaning (grades 1-9).—The Para- graph Meaning test consists of a series of paragraphs, graduated in difficulty, from each of which two or more words have been omitted. The pupil's task is to demonstrate his comprehension of the paragraph by se- lecting the proper word for each omission from the choices that are given. Four other subtests are useful as criteria for the Reading part of the WRAT because, theoretically, they are related to reading. These are the following: Spelling (graaes 1-9).—The Spelling test con- sists of multiple choice questions in which the pupil chooses the correct spelling from among three possible spellings or marks "ng" if the correct spelling is not given. Wora Stuay Skills (graaes 1-4).—The Word Study Skills subtest contains various com- binations of auditory perception of begin- ning and ending sounds, phonics, and phono- grams. For the beginning and ending sounds the pupil must match a word from a multi- ple choice selection matching the beginning or ending sound read by the teacher. The phonics involve selecting a written word which is the same as one he hears read by the teacher, and the phonogram requires matching a word he hears with a rhyming one which he reads. Language (grades 2-5; 7-9)..—This is an exercise in capitalization, punctuation, sen- tence sense, and language usage, with a few additional items of grammar. In all items a correct and an incorrect, or much less acceptable, usage are presented as options. Wora Reaaing (grvaae 1).—Pupils are re- quired to look at a picture and then select the appropriate word from a multiple choice set. This subtest is designed to measure skills to analyze and identify words out of context. The following three tests are used as the principal criterion variables vrelatea to the Arithmetic portion of the WRAT: Avithmetic Concepts and Reasoning (grades 1-9).—The Arithmetic Concepts and Reason- ing tests measure reasoning with problems taken from life experience, with the reading vocabulary being kept much below the prob- lem-solving level being measured. Also tested here is the informational background of pupils and their understanding of the numbers system. Avithmetic Computation (grades 2-9).—The Arithmetic Computation test measures pro- ficiency in computational skills. The tests are multiple choice forms; the response "not given" is included as one of the choices in each question in order to discourage guessing. Avithmetic Application.— This test occurs only at the fourth grade level and is designed to measure application of number concepts to practical situations. Three additional criterion methods of school performance were included which are not airectly 10 related to the predictive variables. These arethe following: Social Stuaies (grades 4-9).—The items in this test primarily measure social studies content or information with approximately equal distribution among history, geography, civics, and social problems. Science (grades 4-9).—This subtest contains about equal proportions of items from the areas of life science, health and safety, elementary physics, and chemistry, with a smaller representation for the earth sciences and conservation. The form used for the second and third grade levels combines the above two topics into a Science and Social Studies Concepts test. Stuay Skills (graaes 5-9).—This subtest measures study tools including reading charts, graphs, and tables; map reading; and using the dictionary. Metropolitan Achievement Test This test contains a total of 11 subtests. Because of the time factors involved, only 7 of the 11 subtests were given routinely, while 1 or more of the remaining tests were given in some of the grade samples. The five criterion variables thought to be most relevant (Reading, Spelling, Language, Mathematical Computation and Concepts, and Mathematical Analysis and Problem Solving) were administered in all instances. One of the Metropolitan subtests, Reading, is a direct criterion for the WRAT Reading test: Reaaing.—This test consists of four reading selections. The student's reading compre- hension is assessed by presenting him with multiple choice questions on content and word meaning. Three other subtests are indirectly relevant as criteria for the WRAT Reading test: Spelling.—This test consists of a number of sentences, each containing one underlined term. The student has to decide whether the term is spelled correctly. Language. —This test covers punctuation and capitalization, recognizing correct word uses, and understanding correct word usage, as well as sentence structure. Language Stuay Skills.— This is a test of the student's ability to use a dictionary and to identify appropriate sources of information. The following two tests serve as criteria for the WRAT Arithmetic Test: Mathematical Computation ana Concepts.— This is a series of arithmetic problems com- parable with those on the WRAT. However, answers are provided in multiple choice form and the procedure of solution may introduce a recognition element. Mathematical Analysis ana Problem Solv- ing.— This is a set of somewhat more com- plex problems expressed in language form. They require the student to identify the prob- lem as well as to select the correct solution from the set of multiple choice answers. The remaining five Metropolitan Achieve- ment subtests are not divectly related to the WRAT: Social Studies Information. —These are mul- tiple ‘choice questions covering history, civics, and geography. Social Stuaies Stuay Skills.—This subtest measures ability to read and interpretmaps, tables, graphs, and charts and also assesses the student's ability to draw inferences from such data. Social Stuaies Vocabulary.—This is a multi- ple choice test of the student's knowledge of terms (taken from newspapers, magazines, and school publications) relating to social science studies material encountered in and out of the classroom. Scientific Concepts ana Unaerstandaing.—This is a measure of the student's science vocab- ulary and of his comprehension of printed scientific material of the kind covered in high school science courses. Science Information.—This consists of mul- tiple choice questions covering a broad area of the physical and biological sciences. EXAMINATION PROCEDURE The criterion achievement battery was ad- ministered by classroom teachers in the conven- tional manner in order to replicate the normal school use of achievement tests. To gain further assurance of normal administration, the services of the public school testing director were ob- tained to direct the group achievement test admin- istration. In some instances the achievement tests were administered in a single day, but at other times 2 days were required. In order to replicate the examination pro- cedure used on the Health Examination Survey, the WRAT was administered individually. Exam- iners were classroom teachers from the partic- ipating schools who had been specially trained in WRAT administration. Although, for convenience and economy, children were examined by class- room teachers from their own schools, inno case was a child examined by his own teacher. Each sample child was given the WRAT during one of two programed times (1) during the hour before the start of the day's classes or (2) during the hour immediately after the end of the day's classes. Children were randomly distributed between these two testing times. A systematic surveillance of the Reading test was effected by tape recording selected testing sessions. Table 12 gives the number of examiners used in each grade for the Monongalia County samples and the control samples. Almost all examiners gave WRAT's to children in all grade levels of the school in which they served as examiners. Dif- ferences innumbers of examiners at the secondary school level between the Monongalia County and control samples occurred because only teachers were used as examiners in the Monongalia County sample while graduate students in psychology were hired to supplement the examining staff in the control samples. The teachers and other examiners were provided with a copy of the administration in- structions lifted verbatim from the WRAT manual (see Appendixes I and II). In addition a training 13 Table 12. Number of examiners administer- ing the WRAT subtests in the Monongalia County samples and in the control sam- ples, by grade Monongalia Grade County Control samples P Number of examiners Grade l-----=-=-=-- 6 4 Grade 2--=--mmmem=- 8 3 Grade 3------------ 8 9 Grade 4----==-=-=---- 5 4 Grade 5---=-===m=-- 5 3 Grade 6------===--- 6 8 Grade 7----=-=---=-- 25 3 Grade J-=rermowmwmns 20 3 Grade 9------=-=--- 24 8 Grade 10-------=--- 23 2 Grade 1ll----=------ 23 3 Grade 12----=------ 23 3 session was conducted for each group of examiners to insure uniform testing procedures. The exam- iners were instructed to serve primarily as recorders of the pupils' responses. They were not expected to do any test scoring. To insure uniformity in scoring and reporting of results, all tests were scored by research personnel. Achievement tests were machine scored directly from the students' answer sheets and then punched on IBM cards for analysis. All WRAT's were scored according to instructions in the manual, and Jastak's norms were used to ob- tain grade level scores. Reliability of scoring was spot checked and is reported in the following section. RELIABILITY OF SCORING PROCEDURE The reliability of scores on the Reading section of the WRAT may have been seriously affected by three sources of technical error. The first of these was the failure of the examiner to record accurately whether the child correctly or incorrectly pronounced a given word. The second source of error was the scorers' varia- bility in interpreting the marks used by the examiners to record the children's performances. A third possible source of error arose from the failure of the scorer to follow instructions to dis- regard correct responses made after 12 consec- utive failures. The first type of error was investigated by checking tape recordings of the Reading exami- nation. Disagreements with the examiners ap- peared to be largely a matter of accepting lo- calisms in pronunciation. The seriousness of this problem is underscored by the fact that for a sample of 30 records, a reviewer who was unfamiliar with local speech patterns obtained a Table 13. Number of scoring errors made in processing WRAT Arithmetic and Reading tests, by sample Arithmetic Reading Rusber of ervors Monongalia | Control | Monongalia | Control sample sample sample sample (N=72) (N=72) (N=72) (N=72) NO error--=-----emom cme eee mmm 71 68 62 70 One error-----=-===m=mee;;;eccce eee. ————— 1 4 7 2 TWO errors---------emecccc eee e ce ee meee - i 2 - Three errors----=----emeece cece eeeeee em x NOTES: Average scoring error per record: Arithmetic=0.035 points; Reading=0.111 points. N—number . 12 Table 14. Means and standard deviations on selected background data, by grade for the Monongalia County elementary school samples Days Occupa- between Age at Grade level ii ahllizy individual time of at time of Grade parent and group group test | group test Mean | S.D. | Mean | S.D. | Mean S.D. Mean S.D. | Mean| S.D. Grade 1 (N=114)------- 4.69] 2.84 -— -—- 5.00 9.26 6.98 0.36] 1.80 0.01 Grade 2 (N=111)------- 4.70 2.45 -—— -—— 1.67 | 8.61 8.08| 0.47] 2.80 0.00 Grade 3 (N=113)------- 4.95] 2.55] 4.23 | 1.00| 44.80 | 4.14) 9.05 0.45| 3.80 0.00 Grade 4 (N=121)------- 4.65] 2.53 4.32 1.05 3.45 6.08 10.17] 0.56 | 4.80 0.00 Grade 5 (N=105)------- 4,721 2.29 4.08 0.97 8.32 5.19] 11.13| 0.67 5.80 0.01 Grade 6 (N=119)------- 5.18 2.32 | 4.41 0.93] 34.77 5.15] 12.00 0.65] 6.80 0.00 Combined grades 2 and 3 (N=224)------ 4.831 2.50 -— --=1 23.42] 22.59 8.57] 0.67] 3.30 0.50 Combined grades 5 and 6 (N=224)------ 4.97 2.32 4.25( 0.96 27.69 18.91 11.59| 0.79] 6.33 0.50 NOTE: S.D.—standard deviation; rho of .60 with local examiner decisions. Prac- tically all disagreements, however, were resolved in favor of the examiners’ scoring when allowance was made for localisms. The other two error sources were investi- N—number. of three boys and three girls from each grade level and rescoring these records. Table 13 gives the frequency distribution of discrepancies and suggests that scoring errors have little effect gated by drawing a random sample of the records on data analysis. Table 15. Means and standard deviations on selected background data, by grade for the elementary school control samples Days Ozcupas between A ge at Grade level (JSlonel, | sbility individual time of | at time of and group group test group test Grade parent tests Mean | S.D. | Mean | S.D. Mean S.D. Mean S.D. | Mean| S.D. Grade 1 (N=103)------ 4.18 2.82 4.73] 0.77] -1.63 8.70 6.95] 0.26] 1.87| 0.05 Grade 2 (N=104)------ 4.061 2.20 | 4.30] 0.90 9.78 13.98| 8.03 0.37| 2.86 0.05 Grade 3 (N=104)------ 4.39 2.85] 4.441 0.70 0.99 3.45] 9.27] 0.55] 3.80| 0.00 Grade 4 (N=100)------ 4.01 2.88 4.82] 1.11|-12.77 6.30 9.97 0.32] 4.88] 0.04 Grade 5 (N=110)------ 4.541 2.50 | 4.25] 1.01 9.95 8.75] 11.03 | 0.40] 5.90 0.00 Grade 6 (N=106)------ 5.12] 2.84 | 5.01 1.20] -4.27 8.17] 12.25] 0.65] 6.80 0.00 Combined grades 2 and 3 (N=208)----- 4.23 2.55 4.37] 0.81] 4.39 11.51 8.65| 0.78] 3.33 0.47 Combined grades 5 and 6 (N=216)----- 4,82 2.69 4.63 1.17 2.97 11.06] 11.63] 0.82] 6.34; 0.45 NOTE: S.D.—standard deviation; N—number. ll. THE ELEMENTARY SCHOOL STUDY BACKGROUND DATA The subjects for the Monongalia County sample for the elementary school study were obtained by the exhaustive testing of pupils in all six grades of three primary grade schools. These included one school in the central resi- dential area, another in a predominantly middle- class area, and a third in a lower-class, semi- rural area. These schools were chosen in order to maximize the likelihood of obtaining a reason- ably representative selection of pupils from the population being studied. Table 14 shows that the desired results were approximated; that is, on ability and socioeconomic indices the pop- ulation was close to, or slightly above, average. Table 14 also includes data on the mean number of days that elapsed between the individual and group tests, the mean age of the students at the time the group test was administered, and the grade level at the time of the group testing. Intervals between individual and group tests for grades 3 and 6 are considerably longer than for the other grades. This may be accounted for by the fact that group test data for these children were obtained from a school-system wide testing program which was conducted ap- proximately a month prior to the data collection for the present study. Table 15 contains similar data on the con- trol samples used in the elementary study. Com- parison of tables 14 and 15 shows that the children in the control sample had parents of slightly lower socioeconomic status but that they averaged slightly higher on group tests of general ability. Mean age at the time of testing for the control samples was within a maximum of 3 months of the Monongalia County samples. The grade levels at the time the group tests were given were very close for the two samples, with a maximum discrepancy of a tenth of a grade level (or 1 month of class time). The matching for the con- trol samples is probably as good as can be hoped for without census-type sampling procedures. Differences in general ability level need, however, be kept in mind when considering discrepancies between the principal and control samples. 14 ADEQUACY OF GRADE LEVEL PLACEMENT Tables 16 and 17 give the means and standard deviations for the entire population of WRAT raw scores and tables 18 and 19 give similar data for the grade level scores. Tables 18 and 19 show that except for the Arithmetic scores of the second and the fifth graders inthe control sample, all subjects obtained WRAT scores somewhat above the actual grade levels of the class at the time the test was administered. WRAT score means for the control and Mo- nongalia samples for the elementary school study differed up to one grade level for the Reading section and up to approximately one-half grade level for the Arithmetic section and for the grade level estimate obtained by combining scores on Reading and Arithmetic. All differences are sig- nificant at the 1-percent level of confidence, ex- cept for the Reading section in grades 4and 5 and for the combined Reading and Arithmetic score in grade 3. The lower performance of the Monongalia first grade sample may have been due to the absence of kindergarten classes. For the other grades, these data imply that the Colorado and Wisconsin samples demonstrated significantly higher skills in Reading than did the Monongalia sample. The Monongalia sample, in turn, signifi- cantly exceeded the California sample on both Reading and Arithmetic and the Colorado sample on Arithmetic alone. Before suggesting that the above results yield positive evidence of overestimation of actual grade placement, attention must again be called to Jastak's contention that the Reading and Arith- metic grade levels, similar to age-scale-derived intelligence quotients, cannot be expected to show systematic increment with grade wise promotion of pupils. To do justice to the test author, it is necessary, therefore, to determine the magnitude of the discrepancies of the findings of this study from the values given in the manual. Unfortunately, Jastak does not provide means and standard devia- tions for the samples on which his tables of norms Table 16. Means and standard deviations on the WRAT, by subtest and grade for the Monongalia County elementary school samples (raw scores) . Arithmetic Arithmetic Reading + Reading Grade Mean S.D. Mean S.D. Mean S.D. Grade l--=--cmmmmo eee 21.31 3.51) 38.57| 9.13] 59.87] 11.79 Grade 2---=--m-mmmmmm meee 26.60 | 2.96 | 50.37 | 8.61 76.97 | 10.30 Grade 3----=--m-mmmmm meee 32.32 | 3.05]61.77| 9.38| 94.08 | 11.11 Grade 4=--=--mmmmmm meee 34.10 | 3.43]166.20| 12.09 | 100.31 | 14.60 Grade 5-=-=--mmmmcm ome 37.31 | 4.57] 70.16 9.28] 107.48 11.81 Grade 6==-=--==mmmm meme mmm 44,00 | 5.88 78.03] 9.10 122.04 | 13.25 Combined grades 2 and 3-=-------c-ceeaoao- 29.49 4.15] 56.12 | 10.66 85.60 13.71 Combined grades 5 and 6-=-=-=-----cccooo-o- 40.87 | 6.27 | 74.34 9.99 | 115.21 | 14.54 NOTE: S.D.—standard deviation. are based. While a test of significance of dis- crepancies is thereby precluded, itis nevertheless possible to make a direct assessment of the mag- nitude of discrepancies by obtaining from Jas- tak's manual the grade level equivalent to the 50th percentile at the mean age of this study's samples. Tables 20 and 21 give the magnitudes of discrepancies from actual grade placement and from Jastak's norms. These tables also contain the t-ratios for the test of significance of mean differences between obtained WRAT grade level scores and actual grade placement. Actual grade Table 17. Means and standard deviations on the WRAT, by subtest and grade for the elementary school control samples (raw scores) : 3 Arithmetic Arithmetic Reading + Reading Grade Mean S.D. | Mean S.D. Mean S.D. Grade l--==-cmmcc meme 24.68 | 2.09 | 44.50 8.54 69.17 9.95 Grade 2-===- mmc 24,70 | 2.30] 47.69 9.87 72.30 11.17 Grade 3==m-- mmm eam 30.81] 2.15] 63.55] 10.27 94.07 12.03 Grade 4==c----mmcm meee eeee ee 34.11 | 3.44] 71.86] 10.40 | 105.97 12.60 Grade S====ccmmc mma een 36.94 | 4.36 68.31] 11.21 | 105.15 13.94 Grade 6-====- == eee 45.07 | 5.731 79.09 11.77 | 124.16 16.38 Combined grades 2 and 3-------ccemmmcueooo 27.75] 3.78 | 55.62 | 12.82 | 83.18 15.91 Combined grades 5 and 6--------ccmmmccaaan 40.93 | 6.50 73.60 12.69 | 114.48 17.91 NOTE: S.D.—standard deviation. 15 Table 18. Means and standard deviations on the WRAT, by subtest and grade for the Monongalia County elementary school samples (grade level scores) : Arithmetic Arithmetic Reading + Reading Grade Mean| S.D. | Mean | S.D.| Mean | S.D. Grade l-----=------mmemememmmme—emmmmemmmo mmm 1.97| 0.59] 2.06] 0.76 2.02| 0.60 Grade 2--=-====--------mmmmmmmmemeommm——-————e——-o 3.02( 0.69] 3.08| 1.03| 3.05 0.72 Grade 3-----=-----m-emmmmmmeeocmmmeoo——-ooo——m 4,491 0.72 4.74] 1.63| 4.62| 1.02 Grade 4-----==-=----mmmmommememmmmmmo mmo mmm 4.917 0.771 5.70) 2.30¢ 5.30 1.43 Grade 5--=-===-------m-mmmmmmmeee meme 5.62| 1.04] 6.33| 2.00] 5.97 1.29 Grade b----===-------mmmmemmmmommmmm—-—---—o m= 7.60] 2.01] 8.24 2.44] 7.92 1.93 Combined grades 2 and 3-=-=m-—rm=im=camwmm mm m= em= 3.76 1.02] 3.92] 1.60] 3.84 1.18 Combined grades 5 and 6--------------==---------- 6.67] 1.91] 7.34 2.44] 7.01 1.92 NOTE: S.D.—standard deviation. placement at the time the WRAT was administered is obtained by adjusting the grade level at the time the group test was taken by the average time elapsed between the individual and group tests. Results of these comparisons suggest that the WRAT tends to overestimate grade level even when Jastak's norms are used to adjust the actual grade level estimate. The adjusted method also yields WRAT overestimates of grade level except for the second grade control sample and for the fourth and fifth grade arithmetic scores on both the principal and the control samples. Since all but the second and fourth grade control samples are slightly above average in Table 19. Means and standard deviations on the WRAT, by subtest and grade for the elementary school control samples (grade level scores) : Arithmeti Arithmetic Reading we Grade Mean | S.D. | Mean | S.D. | Mean | S.D. Grade l--===-----m-mmemmmemmmo omen msm momen mmm 2.56] 0.38] 2.53| 0.81 2.54 | 0.53 Grade 2--=--==-=--mmmmemmemme mee m mm me smo ————— 2.59] 0.48] 2.87| 1.05] 2.73 | 0.66 Grade 3----=-=---m---mmmmmmmmee——me-—-———ooo—o—-- 4.13] 0.59) 5.10 | 1.87] 4.61 1.12 Grade 4-===-==m=-----emmmmmmmmmeem--—-mmoo———m——— = 4,921 0.77) 6.76 | 2.27 5.84 1.40 Grade 5===--===-------mmemmmoe—m--o-o--—-o ooo 5.53] 0.98 6.09 2.25| 5.81 1.46 Grade 6-------==----emmmmeme meme mm mmm ee ———————— 7.87] 1.89] 8.69 2.60] 8.28 2,02 Combined grades 2 and 3-----=---=-=-=-=--=-----=----= 3.36] 0.94] 3.99 1.88] 3.67 1.32 Combined grades 5 And B-==m~— =m mu mmm ww 6.68) 1.90] 7.36] 2.76 7.02 | 2.15 NOTE: S.D.—standard deviation. 16 Table 20. Discrepancies between observed WRAT grade level scores,actual grade level, and Jastak's age norms, by subtest and grade for the Monongalia County elementary school samples Discrepancy from actual grade level Discrepancy from Jastak's age norms Arithmetic Grade Arithmetic Reading + Arith i Reading rithmetic Arithmetic | Reading + Reading D t D t D t Grade l-=-===-e-ea--- +0.15| 12,68 | 40.24 13.38 +0.20 | '3.51 -0.,05 -0.16 -0.20 Grade 2===----c--c-a- 40.21 3.18 | 40.27 12.75 +0.24 13.48 =-0,29 -0.23 -0.26 Grade 3-===-=---c-aa-- +0.54 | 17.94 | +0.79 5.161 +0,57 5.94 +0.44 +0.69 +0.57 Grade 4===--=-ec-c-an +0.10 1.43 40.79 | 3.76 | +0.49 3.77 -0.10 +0.69 +0.29 Grade S5-====----c---- -0.21 12.08 +0. 50 12.56 +0.14 (LLL -0.11 +0.10 -0.26 Grade 6=====----cc--- +0.68 3.70 | +1.32| "6.14 +1.00 5.653 +0.33 +0,97 +0.65 Combined grades 1 i t 2 and 3-==-=-------- +0.38 7.60 | +0.54 5.74 +0.46 8.21 +0.08 +0.24 +0.16 Combined grades 1 1 5 and f=====---o-n-- +0.13 1.14 140.92 | '6.01| +0.59 5.09 -0.17 +0.50 +0.17 significant at the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance of the difference between means. Table 21. Discrepancies between observed WRAT grade level scores, actual grade level, and Jastak's age norms, by subtest and grade, for the elementary school control samples Discrepancy from actual grade level Discrepancy from Jastak's age norms Arithmetic Grade Arithmetic Reading + Arithmets Reading x met.Lc Arithmetic | Reading + Reading D t D t D t Grade l~~=m=nmwmmms +0.69| '18.65| +0.66| '8.25| +0.67| '12.89 +0.74| 40.71 40.72 Grade 2--=m==-muu-- 0.301 16.52f -0.02| 0.20] -0.16 2.54 -0.51 -0.23 -0.37 Grade 3--m====uu--- +0.33| 15.69] +1.30| '7.10| +0.81| 17.36 +0.08 +1.05 +0.76 Grade f-——mwmemmmmmm +0.01 0.13] +1.85| !8.15| +0.93| 4.10 +0.07 +1.91 +0.99 Grade 5-=======uu-- 0.401 14.30 +0.13| 0.61| +0.12 0.86 -0.37 +0.16 40.15 Grade 6-======uuu-- +1.08( °5.87| +1.90| '7.51| +1.49| 17.60 +0.67 +1.49 +1.08 Combined grades 2 and 3--me-mmenm- +0.03 0.71 +0.65| '5,70| 40.33] 14.58 -0.08 +0.54 +0.22 Combined grades y { i 5 and 6-=========- +0.34 3.01) +1.02| '5.83| +0.68 5.67 +0.30 +0.98 +0.64 Isignificant at the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance of the difference between means. 17 Table 22. Means, standard deviations, and discrepancies from actual grade level on the Stanford Achievement Test, Form Primary I, by sample and subtest for grade 1 Subtest Mean| S.D. D t Monongalia County sample (N=114) Word Reading--| 1.80| 0.58 | 0.00 0.00 Paragraph Meaning------ 1.91] 0.57 i +0.11 12.04 Vocabulary----| 2.28| 1.03] +0.48 2-00 Spelling-=-=--=-- 1.95 0.54 | +0.15 3.00 Word Study { Skills------- 2.16| 0.95] +0.36 4.04 Arithmetic----| 2.08] 0.611 +0.211 13.68 Wisconsin control sample (N=103) Word Reading--| 2.17] 0.52] +0.30| '5.88 Paragraph 1 Meaning------ 2.15( 0.59 | +0.28 4.83 Vocabulary----| 2.24] 0.68 | +0.37 | !5.44 Spelling------ 2.29] 0.66 | +0.42 | '6.56 Word Study y Skills======-=- 2,60 1.00 | +0.73 7.37 Arithmetic----| 2.12] 0.48 | +0.25| !5.21 Igignificant at the l-percent level of confidence. NOTE: N—number; S.D.—standard devi- ation; D—algebraic difference; t—t-test of the significance of the difference be- tween means. general ability, it is conceivable that the latter variable may account for the overestimation that occurred in the present study. Comparison of grade level estimates from the WRAT and the criterion measures will further bear on the issue of adequate grade placement and will be examined in one of the following sections. PERFORMANCE ON THE SAT Means and standard deviations for the grade scores on the Stanford Achievement Test are reported in tables 22 through 27. Data have been tabulated separately for Primary Form I, used in the first grade; Primary Form II, used in the 18 second and third grades; Intermediate Form I, used in the fourth grade; and Intermediate Form J, administered to the fifth and sixth grade children. In every case, differences from actual grade level at time of testadministration have been computed, and these, along with the significance and magni- tude of the differences, are given in the above- mentioned tables. Our data suggest that the samples used for the elementary school study perform at, or slightly above, the normed averages on the group achievement measures. There are some minor discrepancies between the Monongalia County and control samples, with the former gen- erally scoring slightly higher in all but the first grade samples. Significantly higher than average perform- ance was shown by all of the first graders on Vocabulary, Spelling, Word Study Skills, and Arithmetic; the Wisconsin control sample also scored higher on Word Reading and Paragraph Meaning. The second and third graders in the Monongalia County sample scored significantly above average on Science and Social Studies Concepts, and the third graders were also above average on Arithmetic Concepts. But the Cali- fornia second grade control sample was below average on all SAT variables, while the third grade (Colorado) control sample exceeded the norms for Word Meaning, Science and Social Studies Concepts, Spelling, Word Study Skills, and Language. Performance of the fourth grade samples was at the normative level except for above average performance on Social Studies by the principal Monongalia sample and significant below average performance on Arithmetic Computation by the Wisconsin Control sample. In grades 5 and 6 the Monongalia County samples attained above grade level means for Spelling, Social Studies, Science, and Study Skills, while the fifth grade sample had below average performance on Arithmetic Computation. The California fifth grade sample scored below average on all SAT subtests, while the Colorado sixth grade sample was significantly above nor- mative levels on Paragraph Meaning, Spelling, Language, and Study Skills. Most of the significant discrepancies of the SAT from actual grade level ranged from one- Table 23. Means, standard deviations, and discrepancies from actual grade level on the Stanford Achievement Test, Primary Form II, by subtest for the Monongalia County samples, grades 2 and 3 Grade 2 Grade 3 Combined grades 2 and 3 (N=111) (N=113) (N=224) Subtest Mean | S.D. D t Mean | S.D. D t Mean | S.D. D Et Word Meaning------ 2.73 0.75 -0.07 0.99] 3.85] 1.12 | +0.05 0.47] 3.30 | L.11 0.00 0.00 Paragraph Meaning-| 2.87| 0.80 | +0.07 | 0.92 3.95] 1.11 | +0.25 | 2.45| 3.42 | 1.11 | +0.12 1.82 Science and Social ' | Studies Concepts-| 3.11| 1.15| +0.31 2.84 4.32] 1.21 | 40.52 4.56] 3.72 | 1.32 | 40.42 5.32 Spelling---------- 2.731 0.79 -0.07 0.93] 3.97 | L.14| +0.17 1.57] 3.36 | 1.16 | +0.06 0.91 Word Study Skills-| 3.02| 1.46 | +0.22 1.58 4.07 | 1.60 | +0.27 1.80 3.55] 1.62 | +0.25 2.45 Language=-==-=--=-=--- 2.83] 1.04] 40.03 0.30] 4.00 L.16| +0.20 1.83] 3.42] 1.25| +0.12 1.64 Arithmetic Computation------ 2,72) 0.61 -0.08 1.38] 3.72] 0.65| -0.08 .31] 3.22| 0.81| -0.08 1.90 Arithmetic 1 Concepts====-=-=-=- 2,75] 0.92] -0.05 0.57] 4.20| 1.20 | 40.40 3.51 3.48 | 1.29 | 40.18 2.47 'Significant at the l-percent level of confidence. NOTE: N—number; S.D.—standard deviation; D—algebraic difference; cance of the difference between means. fourth to three-fourths of a grade level. The discrepancies from actual grade level in the cri- terion must, of course, be considered in evalu- ating the seriousness of the deviations of the WRAT from actual grade level. Appropriate analyses of this complicating problem are re- ported following the discussion of the WRAT's relation to the criterion measures. Table 24. grades 2 and 3 Means, standard deviations, and discrepancies Achievement Test, Primary Form II, by subtest for the t—t-test of the signifi- RELATION OF THE WRAT TO THE SAT Validity coefficients describing the relation of the WRAT and the Stanford Achievement Test have been grouped together for all grades to permit easier comparison. The reader must be reminded again that, due to the age level of from actual grade level on the Stanford California and Colorado control samples, Grade 2 Grade 3 Combined grades 2 and 3 (N=104) (N=104) (N=208) Subtest Mean | S.D. D t Mean | S.D. D £ Mean | S.D. D t Word Meaning------ 2.36| 0.60 | -0.50 | 8.62] 4.15| 1.31 40.35 | '2.74] 3.26 1.35( -0,07 0.92 Paragraph Meaning-| 2.40 | 0.66 | -0.46| '7.19| 4.10] 1.34| 40.25 1.91] 3.25 1.36 -0.08 1.03 Science and Social : Studies Concepts-| 2.62| 0.73 | -0.44| '6.11| 4.31] 1.47| +0.51 3.54] 3.46 | 1.44] 40.13 1.53 Spelling-=--=------ 2,58] 0.87 -0.28 | 13,33] 4,22 1.18| +0.42| '3.62| 3.40] 1.32] +0.07 0.93 Word Study Skills-| 2.40 0.95| -0.46 | !4,95| 4.61| 1.85| +0.81| '4.48| 3.50 | 1.84| +0.17 1:53 Language-==------= 2.54] 0.64] -0.42| '6.77| 4.26 | 1.54| 40.46 | '3.05] 3.40 1.46 | +0.07 0.81 Arithmetic . : Computation-=-=---- 2.30] 0.50 -0.56| 11.67] 3.94| 0.77| +0.14| 1.87 3.12] 1.04| -0.21 4.20 Arithmetic 1 7 Concepts=====-==-=- 2,30) 0.69 -0.56| '8.36| 3.97| 1.32| +0.17| 1.32 3.13| 1.34] -0.20 2.60 'significant at the l-percent level of confidence. NOTE: N—~number; S.D.--standard deviation; cance of the difference between means. D—algebraic difference; t—t-test of the signifi- 19 Table 25. Means, standard deviations, and discrepancies from actual grade level on the Stanford Achievement Test, Intermediate Form I, by sample and subtest for grade 4 Monongalia County sample | Wisconsin control sample (N=121) (N=100) Subtest Mean | S.D. D Mean | S.D. D t Word Meaning--=-=-==-=-c-m=e-uc---= 4,99 | 1.47 | 40.19 1.42] 5.02 | 1.52 | +0.14 0.92 Fansgraph Meaning-=-----======-=- 5.16 | 1.89 | +0.36 2.09] 5.14 | 1.83] +0.26 1.43 Spelling------ccrmcccccnccncnnnn- 5.00 | 1.69 | +0.20 1.30) 4.98 | 1.54 | +0.10 0.65 Word Study Skills-==-==--==-c--=- 4.61 (1.87 -0.19 1.12] 5.25] 1.88 | +0.37 1.97 Language-===-====-c--==cccca-o==- 4,91 1.92|+0.11| 0.63] 4.61 | 1.74 -0.27 11.35 Arithmetic Computation----===---- 4.86 | 0.93 | +0.06 0.71] 4.16 [0.69 | -0.62 8.98 Arithmetic Concepts--=-==-===----- 5.03 | 1.49 | +0.23 1.70 | 5.14 | L.54 | +0.26 1.69 Arithmetic Applications---------- 4.84 | 1.54 | 40.04 0.29 4.83 [1.35] -0.05 0.37 Social Studies----------ce=meeo-- 5.34 | 1.46 | 40.54 | "4.06 | 5.15 | L.59 | +0.27 1.71 Science-=======cemmmcccccncnna——- 5.14 | 1.69 | 40.34 2.211 5.21 11.72 +0.33 1:93 lgjgnificant at the l-percent level of confidence. NOTE: N—number; S.D.—standard deviation; D—algebraic difference; t—t-test of the significance of the difference between means. the children involved, not all criterion measures are available at all ages. Dashes in the columns of tables indicate such missing data. Validity coefficients are given separately for each grade and for the combined second and third grades and combined fourth and fifth grades. Table 28 lists the appropriate validity measures relating the SAT grade scores to the grade level scores on the WRAT Arithmetic section. Values for the relationships with the most pertinent criteria have been boxed. Coefficients for the individual grade samples (for Arithmetic Concepts and Table 26. Means, standard deviations, and discrepancies from actual grade level on the Stanford Achievement Test, Intermediate Form J, by subtest for the Monongalia County samples, grades 5 and 6 Grade 5 Grade 6 Combined grades 5 and 6 (N=105) (N=119) (N=224) Subtest Mean | S.D. D t Mean | S.D. D t Mean | S.D. D t Paragraph Meaning-| 6.13 | 1.71 |+0.33 1.98 | 7.23 2.00 |+0.43| 2.35] 6.71 | 1.95|+0.38 13,04 Word Meaning------ 5.96 | 1.53 | +0.16 1.07 ]7.19] 1.65 |+0.39 2.58] 6.61 | 1.71] +0.29 2.71 Spelling-====-===- 6.23 | 1.46 | +0.43 | 13,03 | 7.45] 1.77 | +0.65| 14.01] 6.88 | 1.74 | +0.55 15,05 Language=-========= 5.85 (2.08 |+0.05| 0.25] 7.17 | 2.23 | +0.37 1.81] 6.55] 2.26 | 40.22 1.52 Arithmetic Reasoning-------- 5.56 | 1.20 | 0.24 | 2.05 | 7.04 | 1.48 | +0.24 1.76] 6.35] 1.55 (+0.02 0.22 Arithmetic ‘ 1 Computation-=----- 5.37 | 0.97 | -0.43 | '4.53 | 6.81 | 1.16 |+0.01| 0.09] 6.14 | 1.30 | -0.19 2.60 Social Studies----| 6.35 | 1.54 | +0.55 13,677.51 |1.75|+0.71 14.44 6.97 | 1.75] +0.64 15,82 Science----==-=---- 6.57 | 1.99 | 40.77 | 13.97 | 7.40 | 2.04 | +0.60 | 3.21} 7.01 | 2.06 | +0.68 15.04 Study Skills------ 6.48 | 1.95 | +0.68 3.58 17.90 | 2.13 | +1.10 5.64] 7.23 (2.17 |+0.90 6.52 'significant at the l-percent level of confidence. D—algebraic difference; t—t-test of the signifi- NOTE: N—number; S.D.—standard deviation; cance of the difference between means. 20 Table 27. Means, standard deviations, and discrepancies Achievement Test, Intermediate Form J, samples, grades 5 and 6 by subtest for from actual grade level on the Stanford the California and Colorado control Grade 5 Grade 6 Combined grades 5 and 6 (N=110) (N=106) (N=216) Subtest Mean | S.D. D t Mean | S.D. D t Mean | S.D. D t Paragraph Meaning-| 5.42 | 1.69 | -0.48 2.98] 7.23] 2.07] +0.53| '2.69| 6.31] 2.09 | -0.03 0.22 Word Meaning------ 5.77{ 1.33] -0.13 1.03] 6.76 | 1.81 | -0.04 0.23] 6.25| 1.66 | -0.08 0.74 Spelling-==-==-==---- 5.68 | 1.59 | -0.22 1.45) 7.75| 1.74 | +0.95| !5.72| 6.70 | 1.96 | +0.36 13.00 Language-=----=----- 5.19] 2.20] -0.71| 3.38] 7.37| 2.16 | +0.57| '2.77| 6.26 | 2.44 | -0.08 0.52 Arithmetic | Reasoning======--- 5.57] 1.28] -0.33 2,701 6.82 | 1.60 | +0.02 0.13] 6.18] 1.57 | -0.16 1.62 Arithmetic Computation-=-=-=--- 5.41 | 1.10 | -0.49 | '4.67| 6.65| 1.15] -0.15 1.36) 6.02 | 1.29 | -0.32 4.10 Social Studies----| 5.84 | 1.38 | -0.06 0.45] 6.91 | 1.78 | +0.11 0.65) 6.37 | 1.68 | -0.03 0.28 Science-=====--=-=-- 5.40 | 1,93| -0.50 | '2.72] 6.79| 2.03| -0.01| 0.05] 6.08 | 2.10 | -0.28 2.06 Study Skills------ 5.58 1.72 -0,32 1.96 | 7.74 | 2.00 | +0.94 | '4.92| 6.64 | 2.15 | +0.30 2.24 Isignificant at the l-percent level of confidence. NOTE: N—number; S.D.—standard deviation; D—algebraic difference; cance of the difference between means. t—t-test of the signifi- Table 28. Validity coefficients describing the relation of the WRAT Arithmetic grade level scores with the grade level scores on the Stanford Achievement Test, by grade, sample, and subtest for the elementary school samples for grades 2 and 3. M—Monongalia County samples; C——control samples. Combined | Combined Grade 1 | Grade 2 | Grade 3 | Grade 4 | Grade 5 | Grade 6 | grades 2 | grades 5 SUBLEst and 3 and 6 M C M C M Cc M C M Cc M C M C M C Word Meaning or Vocabulary-- | .44 | .25| .47 | .35| .33| .59| .60 | .61| .59| .56| .64 | .48 | .59 |.76 .67 +59 Paragraph Meaning=-====== 45) .55 41 50] .51| .61| .59 | .61| .60 | .54 | .58 | .44 | .63|.77 .62 .56 Spelling======== .55| .65| .46 | .45| .36| .54 | .56 | .47 | .46 | .58 | .59 | .64 | .62(.74 +61 wd 2 Word Study Skillg==-====== W341 .51| .21| .43 | .39| .57 | .56 | 54 === | === === | === | 44.73 --- --- Language--=--===-- ~e=| ===] .48| .30| .46| .59| .63 | .59| .61| .42| .65 | .67 | .63[.71 .66 .66 Word Reading---- SL | .55| === | ===] === | === | === | ===] === | mmm | === | mmm | === [=m -—— -—- Arithmetic Concepts and Reasoning==---- 61| .55| .59| .49| .64| .70| .75| .53| .78| .72| .78 | .51| .76|.86 .82 .65 Arithmetic Computation--==| === | === | ,74 | ,59| .64| .61| .72| .68| .63| .69| .74 | .50 | .82|.79 +78 .67 Arithmetic Application==== | === | === | === | === | === | ===] 70 | .63| === | === === | === | === [=== --- -— Social Studies--| -=-- | === «70 | «8371 +453| «37; «359 | «38 .60 32 Sctoncgsremmmmns So| cof 63.28.39] .52) “5g | '55| i59| .59| .53 | .38]| +64|-6% | 55] 153 Study Skills--=-=| === | ===] === | =oc| === | =o=| === | ===] 64 | .70 | 71 | 44 | === |r=- 71 .65 NOTES: Social Studies and Science were combined in a single subtest in the form used 21 Table 29. Validity coefficients the elementary school samples describing the relation with the grade level scores on the Stanford Achievement of the WRAT Reading grade level scores Test, by grade, sample, and subtest for Combined | Combined Grade 1 Grade 2 Grade 3 Grade 4 | Grade 5 | Grade 6 | grades 2 | grades 5 and 3 and 6 Subtest M Cc M Cc M Cc M Cc M Cc M C M C M C Word Meaning or Vocabulary--| .68 | .41[ .79| .76| .79| .75| .83| .70| .68| .79| .66 | .60| .84| .84 | .72 .71 Paragraph Meaning--=-=--- 871.79] .79| 71} .79| .78) .74| J71| .59| .65| .63| .48]) .84| .85] .65 . 64 Spelling-----=== .76 | .80| .77| .84} .78| .81| .76| .63| .64| .76| .72| .80) .84| .88| .73 .84 Word Study Skillg--emmcnn= 64 77) W721 J72| L71| .78)] .83| J78| === ===] === | ===] .74| .85| --= -—— Language=-=-=--- == | === .67| 54 .72| 72 .77| 71] .67| .63] .62| .74| .77| .79| .68 .75 Word Reading---=| .82 | ,8Ll| ===| ==] =c=| mec] mmm] mem | mon | mec | mee | mee | === | mmm | === -—— Arithmetic Concepts and Reasoning----- 64 | 45] .64 | J53) .55| 71) .62 | J46| 47] .60| J54 | 45] .70| .79] .60 .60 Arithmetic Computation----| === | -==| 37 ,33| .55| .54| .57| .65| 21 | .47 | .41 | .43| .64 | .71 | .47 «57 Arithmetic A eaten | | 2 oo Es ge ws mlm ml em ws Social Studies~-| === | === . . . . . . : Science--------- a= | ===] OY] -21] 3%) WSL Teg les | teu 68] isu | tae] +66] +64] leo | 63 Study Skills-=-=| e=o| coc] won| coc} cce| aca] eau | === | 64 63] 61 | 48 | ==={ ===] .67 .65 NOTES: Social Studies and Science were combined in a single subtest in the form used in grades 2 and 3. M-—Monongalia County samples; C-—control samples. Reasoning, Arithmetic Computation, and Arith- metic Application) range from .59 to .78 for the Monongalia County samples and from .49 to .72 in the control samples. Similar coefficients for the combined grades range from .76 to .82 and from .65 to .86, respectively. Substantial construct validity is supported, at least in the principal sample, by the fact that the most pertinent validity coefficients are higher than the associations with other criterion variables which are unrelated to the arithmetic tasks. There seems to be some progression in increased validity from the first three grades to the higher grades. Most likely this reflects the fact that the WRAT has, of necessity, fewer items to be administered to children in the lower grades, thus producing reduced variability. The validity coefficients describing the rela- tion between the Reading section of the WRAT and 22 the grade scores of the Stanford Achievement Test are reported in table 29, Validities for the most pertinent, criteria (including Word Meaning or Vocabulary, Paragraph Meaning, Spelling, Word Study Skills, Language, and Word Reading) are again boxed in this table. Their values range for the individual grades from .59 to .87 in the Monongalia County samples and from .41 to .84 in the control samples. The combined grade samples yield validity coefficients ranging from .65 to .84 and from .64 to .88, respectively. It is again noted that evidence for construct validity may be inferred from the fact that the language- related subtests of the SAT have higher validities than the arithmetic-related ones when compared with the WRAT Reading section. The validity coefficients describing the re- lationship between the combined Arithmetic and Reading grade level scores from the WRAT and the grade scores from the Stanford Achievement Test are given in table 30. Use of the combined Arithmetic and Reading (A + R) score, as pre- dicted, raises validities for the measures which are not specifically language or numerical ability related, but it does not markedly affect coeffi- cients for the more specific measures. DISCREPANCIES BETWEEN WRAT AND SAT GRADE LEVEL ESTIMATES A set of validity coefficients does no more, of course, than indicate the extent to which the distribution of scores on a predictor variable conforms to, or can be linearly transformed into, a set of scores on a criterion variable. It does not in itself give information on the magnitude of discrepancies in estimating grade levels onthe criterion from the predictor measure. In previous sections the discrepancies of the test scores from actual grade level at time of testing have been examined. These are necessarily related to the question, How closely do the samples conform to national averages? Quite independent thereof, and within this closed system, it is possible to examine the question, How well does the WRAT approximate the grade level estimates on the criterion instrument? Discrepancies between the grade level means for the most pertinent criterion variables and the WRAT Arithmetic section are given in table 31 together with the t-ratios for the significance of these differences. It will be noted that the Arithmetic section of the WRAT significantly overestimates achievement on Arithmetic Con- cepts and Reasoning for the second and sixth grade samples as well as for the two combined samples. Arithmetic Computation, moreover, is overestimated by the WRAT at all levels except for the fourth grade sample. Overestimates of the Table 30. Validity coefficients describing the relation of the WRAT combined Arithmetic and Read- ing grade level scores with the grade level scores on the Stanford Achievement Test, by grade, sample, and subtest for the elementary school samples Combined | Combined Grade 1 |Grade 2 | Grade 3 | Grade 4 | Grade 5 | Grade 6 grades 2 | grades 5 SUbEGSE and 3 and 6 M Cc M Cc M Cc M Cc M C M C M C M C Word Meaning or Vocabulary----- 65 .40 | L79| .73| .74| .77) .83| .70| .77| .79| .75| .61| .82 | .87| .79 .70 Paragraph Meaning-------- 77.79.76 .75| .81| .81| .75| .71| .70| .68 | .70 | .51 .84 | .88 | .72 .67 Spelling------~-- .75| .841 .722| 83} .75| .8%| .77 .63) .,68| .78}| .76| .81; .83 | .89]| .77 +35 Word Study Skills==-===-== 57 .76 | .62] .73] .70| .79]| .82| .78| === | === | === | ===] .69 | .86| --- --- Language=-==-=-=--=-- -==| ==-1.71| .54] .73| .75] .80] .71}| .77 | .63 {| .73 | .79}| .79 | .82] .75 .78 Word Reading-==-| .77| 8l | ===] ===] === | === | === | mom | mom | mom | mmm | mmm | mmm | mmm | == --- Arithmetic Concepts and Reasoning----- .70| 54.74) .60| .66| .77 | .71| .46| .68 | .70 | .75 | .52| .80 | .84| .79 .67 Arithmetic Computation=-=-=--| === | === | 61 | .47| .66| .61| .66 | .65| .41| .59 | .65| .51| .79 | .81 | .69 .66 Arithmetic Application----| ===] === | ===] ===] === ===] .62 52 mmm | mmm | mmm [mmm | mm | mm | --- Social Studies--| ---| --- .70 | 70.75) .73 | .51 .75 .65 SelengE-cannnas ~ee) +06) .26[ .6L| 36] '75| 08, .74| .72 | 62.48] 72-70 65] 63 Study Skills----{| ===] === | -~-] ===] -=~-|{ =~} -~-| ~~~] .76]| .70] .76 | 51] == | ---] .78 +71 NOTES: Social Studies and Science were combined in a single 2 and 3. M—Monongalia County samples; C—control samples. subtest in the form used for grades 23 Table 31. Discrepancies between selected criterion measures from the Stanford Achieve- ment Test and the Arithmetic section of the WRAT, by subtest, sample, and grade for the elementary school samples Arithmetic Concepts and Reasoning Arithmetic Computation Monongalia Monongalia Grade County * | Control | Mout, ! | Conerol samples P samples P D t D t D t D t Brade L===mwemsnnavanaannns -0.13 | 2.60 | +0.44 | 10.73 ... 03 2200 gems OTAdE Rewmmm mem smrm mm sme +0.26 | '3.61| +0.26 | '4.33[+0.29| '6.30| +0.26 5.91 rr I ———————— +0.14 | 1.63] 40.16 1.63 |+0.62| '11.27| 40.19 | 3.11 I EE ———— -0.13 | 1.33| -0.25 1.91 [+0.04| 0.71] +0.73| '12.37 Cade Swwees mmm em wm sme +0.05 | 0.68 | -0.07 | 0.82 |+0.24| 12.86| +0.09| 1.14 BLAS Bums namamin mummies +0.44 | 13.46 | +1.06 6.23 |+0.67| 15.28] +1.23 7.69 Combined grades 2 and 3----| 40.20 | '3.77| 40.22 | '4.40 |+0.44| '11.00| +0.23| !5.23 Combined grades 5 and 6---- | +0.23 | !3.15| +0.50 | '4.95|+0.44| '5.50| +0.66| '6.54 'Significant at the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance of the tween means. arithmetic criteria range up to a maximum of two-thirds of a grade level and are replicated by the control sample in all instances except Arith- metic Computation of grade 5. Similar data for the Reading section of the WRAT are presented in table 32, Here it is obvious that the WRAT overestimates the cri- terion grade levels for all but the Monongalia County first grade sample. WRAT Reading scores exceed significantly the SAT Word Meaning scores for grades 2, 3, 4, and 6. The WRAT Reading score significantly exceeds the SAT scores for Paragraph Meaning and Language at all grade levels and the Spelling score at all but the first and fifth grade levels. It is also significantly higher than Word Study Skills at the third and fourth grade levels and at the second grade in the control sample. The magnitude by which the WRAT Reading section overestimates the cri- terion measures of verbal skills ranges up to a 24 difference be- full grade level and averages at approximately one-half grade levels. Again these findings are uniform for both samples except for Word Mean- ing at the first grade level. THE WRAT SECTIONS Interrelationship Since the WRAT is being used as a brief estimate of school achievement, one must further ask the question whether combining the scores from the Arithmetic and Reading sections would provide a more adequate predictor of grade place- ment or whether either of the two sections might prove to be a sufficient brief estimate of achieve- ment level, The relevant data are presented in table 33, where the intercorrelations be- tween the two sections of the WRAT are listed together with the correlation of each individual Table 32. Discrepancies between selected criterion measures from the Stanford Achievement Test and the Reading section of the WRAT, by subtest, sample, and grade for the elementary school samples Word Meaning or Vocabulary Paragraph Meaning Spelling Monongalia Monongalia Monongalia Grade County Consral County Control County Coneyol samples P samples P samples P D rE D t D t D t D t D Grade l---| -0.24| !3.38| 40.29 '3.62|+0.13| 3.61 | +0.48| '9.80| +0.09| 1.92 +0.24| !5.00 Grade 2---| 40.34| 15,67 | +0.48 | 15.05 | 40.24 | 13,94 | 40.44 | 16.03 | 40.34 1s. 23 | 40.26 14.64 Grade 3---| +0.74 17. 79 | 40.95 L7. 79 | 40.64 16. 74 | 41.00 18.69 +0.62 6. 39 | +0.88 7.86 Grade 4---| +0.70( '5.60| +1.71 10.56 | 40.53 | "3.68 | +1.59 9.88 40.69 | '5.07| +1.75 9.89 Grade 5---| 4+0.34| 2.36 40.29 2.10 +0.17 1. 02 | +0.64 13.90 +0.07 0. 46 | +0.38 11. 71 Grade 6---| 40.93 | 15,57 | +1.94| 19.56 | +0.89| 4.97 | +1.47 6.26 | +0.67 | 4.32 | 40.95 6.13 Combined grades 2 1 1 1 1 1 1 and 3---| +0.54| '9.15| 40.72 | "10.00 | +0.42| “7.00 +0.73| '10.28| +0.48 | "8.28 | +0.58 8.79 Combined grades 5 { 1 1 1 1 1 and 6---| +0.64| 5.61 | +1,11 8.28 | 40.54 | "4.29| +1.05 7.24 40.37 | "3.33 | 40.66 6.23 Word Study Skills Language Grade Monga Control Mogngelie Control samples samples samples samples D t D t D t D t Grade l-------ecccccccoccc ccna -0.08| 1.14 -0.07 1.11 --- -— --- -—— Grade 2---cecceemcccccncncccccacccnanan +0.05| ,0.52| +0.44 15.94 +0. 24 13 04| +0.30 is. 90 Grade 3---cemcccccccmcccc ccc eee +0.52 | (4.48 40.49 4,08 40.59 15. 51| +0.84 "6.51 Grade 4---ceccmccccncccanrccnccncannnan +1.08 | '9.08| +1.47| '10.50| +0.78 '5 82) 42.12 113.25 Grade Semmssnmmawmrnesmmnens wm —— -—— —— ---| 40.45 '2 78| +0.87 La. 75 Grade Gemeeemcemecscennnceansnecannn.— ——- -—— -—— -=-| 40.95 | "4.15| 41.33 17.78 Combined grades 2 and 3-----ececa-ca-- +0.29 3.72 +0.48 6.76 | +0.42 6.05 +0.58 17.25 Combined grades 5 and 6---ecceccceanca-n — —— po —=-| 40.70 | '5.51| 41.10 1g. 73 Significant at the l-percent level of confidence. NOTE: section with the combined Arithmetic and Reading score. It is found here that from one-fourth to one-third of the variance of the two sections is common. The Reading section, moreover, is a superior estimate of the combined score and D—algebraic difference; t—t-test of the significance of the difference between means. would therefore be chosen as the appropriate de- vice if it were deemed necessary to reduce the time allocated by the Health Examination Survey to the achievement measures. This conclusion on the use of parts of the WRAT as a sufficient 25 Table 33. lation of each Correlation of the Arithmetic and Reading sections of the section with the Arithmetic and Reading WRAT and corre- (A + R) composite score, b sample and grade for the elementary school samples (raw scores and grade level Sena) Gl Arithmetic | Reading : ith + R i R Grade Reading with A with A + M C M C M Raw scores Grade l-===-- o-oo momen .67 .61 .82 .73 .98 +99 Grade 2-====- mmm ee eee 46 + 31 .67 .66 .97 .98 Grade 3--=-==mmm meme een 45 +37 .66 .68 .97 .96 Grade f--=-- mmm ee eam .66 .54 .78 le .98 +97 Grade S===== mmm eee .38 +32 .69 .72 .93 .96 Grade 6====- mmm em +55 + 72 .83 .87 .94 .97 Combined grades 2 and 3===------coommmmm oo +65 .75 .81 .84 .97 .98 Combined grades 5 and 6---===-cccmmmmmmm eee - .58 71 .83 +87 .94 +97 Grade level scores Grade l-==-m=m comm mean .58 «37 .86 .78 +92 .96 Grade 2===--- mmm eee +39 43 .75 .70 .90 +95 Grade 3==--=-m meme em 43 +39 .70 74 85 .98 Grade 4==mmm mmo ee eam +63 . 54 .78 +2 +28 +97 Grade S=-=-- memo ee eam .38 .58 .69 .78 +93 .96 Grade 6====- mmm eam +30 .62 .84 .86 +89 +23 Combined grades 2 and 3-=----=--cocommomomaooo +62 «723 +35 .88 .94 +97 Combined grades 5 and 6-==--=----ocommmmmmmma .56 «20 .85 .89 91 «25 NOTE: M—Monongalia County samples; C—control samples. estimator of school achievement is further sup- ported by the data reported in tables 28 to 30 and is discussed in the preceding section. Relation to General Ability and Socioeconomic Status The relation of the WRAT to general ability and to the socioeconomic status of parents’ is of some concern in interpreting these results. Correlations with crude indices for the above 26 variables are therefore given in table 34. Re- lation to parent's occupation ranges from quite low to moderate, and it may be concluded that the WRAT is probably equally suitable for children of different economic backgrounds. Moderate correlations with general ability were found, and they tend to increase with age. While there is relatively little correlation in the first two grades, the relationship increases to the pointofaccount- ing for one-fourth to one-half of the common variance. Table 34. school samples Correlation of the WRAT grade level scores with general ability level and with occupational level of parent, by subtest, sample, and grade for the elementary — ere Occupational level General ability level with: of parent with: Grade : ‘ ‘ Arithmetic v v . Arithmetic Arithmetic | Reading + Reading Arithmetic | Reading + Reading M Cc M C M Cc M C M C M Cc Grade l-==mwm===- ie 371 =~ L33 om +36 .28 10.30] .24 32 vad Grade 2--=m==m==-~ ww .10| ---| .08 .10 21 01 .28]| .07 29 .06 Grade 3----=----~- .34 341.59 | .31 .59 .35 +21 39 31] «alk v2 avn Grade 4==-=--=----=-- .54 56 | 52] W511 .56 «57 «Lb .30| .16| .29 .17 +32 Grade 5---------- 47 +38 1 +35] «93 45 .60 +30 L231] 21 .27 .28 .28 Grade 6-----=-=--- .65 .32| .62]| .58 +d .62 .26 14] L250 .34 +28 +28 Combined grades 2 and 3-- a 19 ===] .21 -—- .22 .19 19 .28| .28 +237 wl? Combined grades 5 and 6-- .57 .59 | .51 | .62 #39 .66 a5 .20| .25| .32 .29 .29 NOTE: M—Monongalia County samples; C—control samples. lll. THE JUNIOR HIGH SCHOOL STUDY BACKGROUND DATA Subjects for the junior high school study in Monongalia County were obtained by the exhaustive testing of all students in the lower three grades of a semirural junior-senior high school and of all students in a suburban junior high school. A semirural suburban junior high school was also used in the three control samples. This particular selection of schools appeared to be the most fea- sible one for obtaining a broad socioeconomic representation. Tables 7 and 9 show distributions which suggest that this objective was generally reached. The summary given in table 35, however, suggests some noteworthy discrepancies in the junior high school samples between the Monongalia County samples and the control samples. The Monongalia samples showed an average socio- economic level quite characteristic for that re- gion but significantly lower than the level found for the control samples. Moreover, while the Monongalia samples were centered at about the national average for intellectual ability, it was found that both the Wisconsin and Colorado samples were above average on intellectual ability. These regional discrepancies must be noted and taken into account in the interpretation of findings for the junior high school samples. Table 35 also gives data on age at time of test and actual grade level. Here the Monongalia and control samples were close to each other. 27 Table 35. Means and standard deviations on selected background data, by grade sample for the junior high school samples and - Days Osgupes Ability between Age at Grade level level of level indivioual time of at time of Grade and sample parent ns Stoup group ‘test | group test Mean | S.D. | Mean | S.D. Mean S.D. Mean | S.D.| Mean | S.D. Total junior high Monongalia County samples--==-==cecau- 3.50 2.57 3.92 0.95 1.211 12.73 14.06 | 1.09| 8.82 | 0.83 Control samples------ 4.95 2.40) 4.58 | 1.01 =-4.31| 5.94) 14.08 0.92] 8.88 | 0.80 Grade 7 Monongalia County samples-===-=c-c-c-- 3.87| 2.65 3.86 | 1.03 8.40 7.091 13.02| 0.66] 7.80 | 0.02 Control samples------ 5.41| 2.25 4.72 0.89| -6.04| 3.82] 13.05| 0.42| 7.85] 0.05 Grade 8 Monongalia County samples-====c=--oo-- 3.12 2.63 3.92| 0.93] 10.73| 5.42] 14.08| 0.78] 8.80 0.01 Control samples------ 4.54 2.24) 4.11 0.89] -9.03| 4.55| 14.06] 0.45] 8.90 | 0.00 Grade 9 Monongalia County samples-====-=ce-aa- 3.48 2.39] 3.97] 0.87) -13.69| 5.68 15.03| 0.66] 9.79| 0.02 Control samples------ 4.92) 2.59) 4.91] 1.06 1.78 | 2.63] 15.05 0.43] 9.80| 0.00 NOTE: S.D.-—standard deviation. ADEQUACY OF GRADE LEVEL PLACEMENT Means and standard deviations for the raw scores on the Wide Range Achievement Test are reported in table 36. Similar data on grade level scores appear in table 37. While the elementary grade study showed generally higher performance for the Monongalia County samples, the reverse is true for the samples under consideration here. Particularly noteworthy for the Reading section of the WRAT are regional differences, which range from one-half of a grade level to more than two grade levels. 28 Comparison of obtained grade level with actual grade level and discrepancies from Jastak's norms as obtained by referring to the values given at the 50th percentile are reported in table 38. The regional discrepancies present some diffi- culties in drawing adequate conclusions. However, it may be noted that both the West Virginia and the California eighth grade samples indicate significant underestimation of reading level. For the seventh and ninth grades, however, reading level is underestimated for the West Virginia samples and overestimated for the control sam- ples. The WRAT Arithmetic section underesti- mates actual grade level throughout except for the ninth grade Colorado control sample, Relating these findings to the evidence on general ability levels in the samples, it may be concluded that the WRAT estimates which are high simply reflect above average intellectual functioning in the respective samples. Thus, it must still be concluded that, in general, the WRAT underestimates actual grade level place- ment at the junior high school level. PERFORMANCE ON THE SAT Means and standard deviations for the grade scores on the Stanford Achievement Test are re- ported in table 39. Mean scores for the Monon- galia County samples were slightly below the ex- pected values, but discrepancies were by no means as extensive as those reported for the WRAT. Mean scores for the Wisconsin and Colorado control samples were also below the expected but were considerably closer to the national norms. The Colorado ninth grade sam- ple, however, tended to equal or exceed the national norms. The Monongalia County samples did better on the Arithmetic subtest of the SAT than on the language-related tests. This appears to be a geographical peculiarity and is not rep- licated for the control samples. RELATION OF THE WRAT TO THE SAT Table 40 lists the validity coefficients de- scribing the relation of the WRAT Arithmetic grade level scores with the grade scores of Ad- vanced Form Km of the Stanford Achievement Test. Coefficients are given for the three grades and for the combined junior high school sample. Coefficients giving relationships to the most pertinent criteria have been boxed. The coeffi- cients for Arithmetic Reasoning and Arithmetic Computation range from .74 to .80 for the Monon- Table 36. Means and standard deviations on the WRAT, by subtest, grade, and sample for the junior high school samples (raw scores) : Arithmetic Arithmetic Reading + Reading Grade and sample Number Mean S.D.| Mean S.D. Mean S.D. Total junior high Monongalia County samples----------- 330 24.35 5.19) 46.87 12.72] 71.371 16.17 Control samples-=---=--ccmcemcmanu-= 327) 26.20 | 6.27) 54.59 | 11.21} 80.81 15.77 Grade 7 Monongalia County samples-=-=---=-===-- 111 | 22.46 | 4.53] 42.94] 11.65 65.40 14.82 Control samples-=---==ccmccmcmoomuno= 104 | 22.47 | 4.16] 55.03| 10.01 77.50 12.43 Grade 8 Monongalia County samples-=-----=----= 101 | 23.89 | 4.59] 45.70) 11.02] 70.09 | 13.06 Control samples==-----=--=cocmeeuou- 109 | 24.80 | 5.46| 48.68 | 11.84] 73.48 15.95 Grade 9 Monongalia County samples-=---------- 118 | 26.52 | 5.46 51.57 | 13.54 78.08 | 17.28 Control samples-------==c=cccmecnn= 114 | 30.96 | 5.60| 59.84 | 8.59] 90.85 | 12.92 NOTE: S.D.—standard deviation. 29 Table 37. Means and standard deviations on the WRAT, by subtest, grade, and sample for the junior high school samples (grade level scores) Grade and sample Arithmetic Arithmetic| Reading + Reading Number Mean | S.D.| Mean | S.D. | Mean | S.D. Total junior high Monongalia County samples------=cccc-c-a- Control samples===-=--c=cocomomonee maa Grade 7 Monongalia County samples====---=----cc-a- Control samples====---comcccmmme ea - Grade 8 Monongalia County samples===-===---coo--- Control samples=-=---=m-o-cmmmoe meee Grade 9 Monongalia County samples--------c-eeo-o- Control samples=====--mcmceomm cee 330) 7.79 | 2.03] 8.36 | 2.90] 8.07 | 2.20 327) 8.62 | 2.61(10.11| 2.73] 9.36 2.38 111} 7.03 | 1.57] 7.41 | 2.60] 7.21 1.87 1041 7.03 | 1.50110.19 | 2.44] 8.61 1.68 101} 7.60 | 1.81] 8.12 | 2.34 | 7.86 1.77 1091 7.98 (| 2.13] 8.68 2.79| 8.33 2.22 118 8.68 | 2.24 9.45 | 3.211 9.07 | 2.43 114 110.68 | 2.50 |11.41 | 2,21 |11.04 | 2.13 NOTE: S.D.-—standard deviation. galia County samples and from .66 to .84 for the control samples. These values are substantial, and construct validity is again suggested since the most per- tinent coefficients relating the Arithmetic section of the WRAT to the Arithmetic Contentare higher than values relating the WRAT to other criterion tests. Validity coefficients describing the relation between the Reading grade level scores from the WRAT and the grade scores on subtests of the Stanford Achievement Testare given in table 41. Coefficients for the most pertinentcriteria (Para- graph Meaning and Word Meaning) range from .57 to .80 and from .47 to .73, respectively. In this instance validity coefficients are almost as high for most other criterion variables with the ex- ception of Study Skills and the Arithmetic tests. Table 42 gives the validity coefficients for the relationship between the combined Arithmetic 30 and Reading grade levels from the WRAT and the grade scores from the Stanford Achievement Test. Validity coefficients in this instance range from .51 to .84 for the Monongalia County samples and from .53 to .79 for the control samples. It may be noted that combined scores will improve prediction for the language-related criterion variables, but not for the number-related crite- rion variables. DISCREPANCIES BETWEEN WRAT AND SAT GRADE LEVEL ESTIMATES Attention must again be given to the question whether the reported underestimate of actual grade level placement noted for performance of the junior high school samples on the WRAT may not be a function of the samples' achievement levels being below their actual grade placement, Discrepancies have therefore been computed be- Table 38. Discrepancies between observed WRAT grade level scores,actual grade level, and Jastak's age norms, by subtest, grade, and sample for the junior high school samples Discrepancy from Discrepancy from actual grade level Jastak's age norms Arithmetic Grade and sample Arithmetic Reading + Reading Arithmetic Arithmetic | Reading + Reading D t D D Lt Total junior high Monongalia County 1 \ samples~==mmmm——=---- -1.03 | 9.81| -0.46 '3.03 -0.75| 6.64 -0.81 -0.24 -0.53 Control samples=-=----- -0.27 | 2.18| +1.22| '8.13|+0.47| '3.88 +0.02 +1,51 +0.76 Grade 7 Monongalia County g sampleg--========---~ -0.75 | 5.03 -0.37| 1.50 | -0.57| 13.22 -0.82 -0.44 -0.64 Control samples------ -0.84 | '5.71| +2.32| 19.67 | 40.74 | 14.48 -0.82 +2.34 +0.76 Grade 8 Monongalia County y | samples-======c----~ -1.17 16.50 -0.65| '2.79 | -0.91| !5.17 -1.00 -0.48 -0.74 Control samples==----- -0.94 | '4.63| -0.24| 0.90] -0.59| '2.78 -0.67 +0.03 -0.32 Grade 9 Monongalia County samples-=====-=-=----= -1.15 5,58 -0,38 1.29 | -0.76 | 3.41 -0,.52 +0.25 -0.13 Control samples------ +0.88 | '3.76| +1.61| !7. +1.24| 6.23 +1.48 +2.21 +1.84 !significant at the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance tween group means for the WRAT sections and the most appropriate criterion measures from the Stanford Achievement Test. Table 43 lists the discrepancies between means on the WRAT Arithmetic section and the Arithmetic Concepts and Reasoning and the Arithmetic Computation tests on the SAT. While there are systematic trends for all but the ninth grade control samples in the direction of under- estimation of the SAT by the WRAT, only a few of these discrepancies reach significance at the 1-percent level of confidence. The only individual grade sample reaching significant levels of under- estimation is the seventh grade Wisconsin control sample. However, when the total Monongalia County junior high school sample is combined, of the difference between means. significant underestimation is found for both criterion measures. But the magnitude of the underestimation averages to two-tenths ofa grade level, and it may therefore be concluded that the use of Jastak's arithmetic grade level norms at the junior high school level will result in under- estimation of actual grade placement but rel- atively accurate placement in terms of the stu- dents' achievement as measured on the Stanford Achievement Test. A rather different story emerges for the Reading section of the WRAT. Relevant data on the discrepancies and their significance are re- ported in table 44. It will be noted that the WRAT Reading section systematically tends to over- estimate the SAT performance. The extent of 31 Table 39. Means and standard deviations on the Stanford Achievement Test, by grade, subtest, and sample for the junior high school samples (grade level scores) Total junior high Grade 7 Grade 8 Grade 9 Subtest and sample Mean | S.D, Mean | S,D.,| Mean | S,D.| Mean S.D. Paragraph Meaning Monongalia County sampleS======--- 7.93 2:31 7.25] 2,23] 7.84 | 2.34 8.65 2.14 Control samples---=--cecemccmcccanon 8.78 2.39 8.19 | 2.24) 7.96 | 2.47] 10.09 1.80 Word Meaning Monongalia County sampleS===-w=---- 8.06 2.36 7.11} 2.16) 8.01} 2.13 9.00 2.34 Control samplesee--cececmmocmaaoao 9.57 2.24| 8.88 .20| 8.89 2.32] 10.08| 1.52 Spelling Monongalia County samples--------- -— -—— ==-| ==--| 7.90] 1.99| 8.90] 2.34 Control samples-==-c--ceccamcacanaa 8.71 | 2.10| 8.34] 1.93] 8.23 07] 9.51 2.04 Language Monongalia County samples=----e--- 7.57 2.99 7.14 | 3,04f 6.57 | 2.61| 8.83| 2.81 Control sampleS-=--eemcmcccccaanan 8.90 2.46) 7.82| 2.32] 8.63 | 2.56] 10.15| 1.86 Arithmetic Reasoning Monongalia County sampleS-e=-=e-=- 8.02 | 2,10 7.14 1.79( 7.92 | 1.90 8.94| 2.15 Control sampleS=-==--ececececmcaancan 8.74 | 2.05|| 7.74| 1.60 8.11 | 1.91 10.27| 1.62 Arithmetic Computation Monongalia County samples=-=e--a-- 7.98 | 1.84 7.23] 1.54] 7.75| 1.56) 8.89] 1.94 Control sampleS-=-=eecececmcncccaaan 8.54 2.10 7.13 1.20| 8.00 | 1.74] 10.35 1.74 Social Studies Monongalia County samples-=ee-w--- 7.70 2.19 6.77] 1.83] 7.62 | 1.94 8.65 2,29 Control samples------eccccccmnaannn 8.71 2.28 8.14 1.98 8.04 | 2.30 9.88 2.07 Science Monongalia County samples===-e==-- 8.35 2.55 7.32 2.48] 8.32 2.27] 9.36] 2.44 Control samples---m-eecemccacaeanan 9.16 | 2.55|| 8.99| 2.30] 8.11 2.69] 10.31| 2.12 Study Skills Monongalia County samples=-=----=-- 7.48 | 2.39| 6.75| 2.20| 7.29] 2.34] 8.33] 2.34 Control samplesS=---memccmemccaanann 8.73 2.45 8.41 .39| 7.65] 2.43 10.06 1.83 NOTE: S.D,—standard deviation. 32 Table 40. Validity coefficients describing the relation of the WRAT Arithmetic rade level scores with the grade level scores on Advanced Form Km of the Stanford Achieve- ment Test, by grade, sample, and subtest for the junior high school samples Total junior Grade 7 | Grade 8 | Grade 9 Subtest high M Cc M Cc M Cc M C Paragraph Meaning-----cceccccccccccccccennnax .59| .66 || .67| .53| .52| .67| .53 | .63 Word Meaning-=-ec-ecccccenmneccacccccecnnna= .58| .64 || .65| .56 | 44 | J58| .51 .61 Spellingee-=ceeeccccncmeccenanaax Seememmm——- -==| 56 || -=--| .49]| .40 | .58] .57 | .50 Language===-==s-eemcecccecccnecmcecneaen———— .58| .66 || .68| .52| .46| .60| .54 | .64 Arithmetic Reasoninge----ecececccccececmccaaanaa-a .80| .80 | .78] .66| .79| .79| .75| .71 Arithmetic Computatione--eececccaccaccoaana-- 79] 84 f 74.69] .78| .78] .76 | .74 Social Studies------cecccccocmemreencnnnana 57] 64) .61| .56| .42| .58] .51 +39 Science------cecccmcccccmcnmm meee mmm .60| .57 0060 471.50) .62] 53] +52 Study SkillS==-ee--ccccccmmem mmm c emcee naam .67]| .63 | .74| .51| .64| .68| .58 | .58 NOTE: M—Monongalia County samples; C-—control samples. Table 41, Validity coefficients describing the relation of the WRAT Reading grade level scores with the grade level scores on Advanced Form Km of the Stanford Achievement Test, by grade, sample, and subtest for the junior high school samples Total junior Grade 7 Grade 8 Grade 9 Subtest high M Cc M C M Cc M Cc Paragraph Meaninge---=-ecececceccccccaccaccnn .68| .64 JI1 | 47) 57 .70| .69 .56 Word Meaning-==-c-ceccccccccccccmacncncnna~n .78] .70 .80| .62| .72| .73| .75 .64 Spellinge-ecmececccaccrcercccmcccccncannnun" --=| .73 --=] .66] .72| .81| .82 .67 Language-====cccecmcccaccccncaccecemene—e——— .70| .63 .73| 52) .62 | .72] .71 +63 Arithmetic Reasoning----=-cececccccecccncnnn 64] .60 62] .54| 51 | .61| .64 +3 Arithmetic Computatione--=-ecececcccecccccancnnn .64| .53 .65| .46| 40 .57| .65 .50 Social StudieS------c-ccccccoccrccccnmnnnnna .64| 61 .70| .59| 44 .67] .63 42 Sciencees=-emecmecccccccac ccm eee e .70| .64 .76 | .55| .55| .66| .68 48 Study SkillS-=-=c-ccmccccccmmmcnmcccncnnnn—— .65] .58 .62| .43| .60| .63| .63 .38 NOTE: M-—Monongalia County samples; C-——control samples. 33 Table 42. Validity coefficients metic and Reading grade level scores with describing the relation of the WRAT combined Arith- the grade level scores on Advanced Form Km of the Stanford Achievement Test, by grade, sample, and subtest for the junior high school samples Total junior Grade 7 Grade 8 Grade 9 Subtest high M Cc M C M C M C Paragraph Meaning-=--eeeeecccecccecmcccamann 72.73 .781 .58| .65| .76| .70 .66 Word Meaning-=-=---ccecmmccccccaccccccccnanax .78 | .75 .84| .70| .70] .74] .73 .69 Spellingeseececccccce mcm cccmemeeceeaaen --=| 72) =---| .69]| .68| .79| .80| .64 Language--===eeeemmecceme meee e eee 2131 #72 .79] .61] .65] 74] .72 .70 Arithmetic Reasoning------eeececemceccccceaeax 790 W791 .76| .69) 74.76) .77| .68 Arithmetic Computatione--=e-eeeeccccmceceaans 78 | «77 77] 64 66 .73] .78 . 69 Social Studies--=-ecmccmmcmccmcc me caeea .69 | .70 .75| .68| .51| .70]| .66 .56 Science=smeeemmcm cme 74 | .68 .81| .61] .62( .71] .69 +55 Study Skillse--em-omcccoo mmm c cme JJ4 | .68 75) 54.73) 72] .68 «53 NOTE: M-—Monongalia County samples; C-—control samples. the overestimate depends also on the criterion measure involved and shows decided regional differences. Thus, the WRAT Reading section significantly overestimates scores on all cri- terion measures for the Wisconsinand Colorado samples. Significant overestimates for the Mo- nongalia County samples occur only for the eighth grade on Language, for the ninth grade on Paragraph Meaning, and for the combined junior high samples on both Language and Para- graph Meaning. Finally, the California (eighth grade) sample is overestimated on the Language subtest only. In summary, it appears that there are sub- stantial discrepancies between reading skill and actual grade placement in the junior high school level, although a definite relationship does exist. As a consequence, we find the apparent paradox that the WRAT Reading test in some instances will underestimate actual grade placement while overestimating language-related achievement as measured by the Stanford Achievement Test. In a situation such as this, the test author obviously faces the dilemma as to whether tokeyhis meas- ures to grade placement or to independent achieve- 34 ment indices. In the case of the norms for the WRAT for the pupils 13-15 years old, the test author apparently has achieved a compromise between these two alternatives. THE WRAT SECTIONS Interrelationship The issue of the interrelation of the two parts of the WRAT used in the Health Examination Sur- vey and the question of the advisability of com- bining these scores has also been investigated for the junior high school samples. Table 45 gives the correlations between the two sections and their relation to the combined score for each grade and the total junior high school samples. Again, sub- stantial correlation between the two sections is noted, and as in the elementary school study, it appears that the Reading section correlates most highly with the combined score. Attention is again called to table 41, which showed that the Reading section of the WRAT predicted per- formance on the Arithmetic criterion variables reasonably well, although not quite as well as did Table 43. Discrepancies between selected criterion measures from the Stanford Achieve- ment Test and the Arithmetic section of the WRAT, the junior high school samples by subtest, grade, and sample for Arithmetic i Concepts and Sxignnesis Reasoning omputation Grade and sample D t D t Total junior high Monongalia County samples--=---=----====c===-==----------=- -0.23[13.19 -0.19 12.75 Control samples--------===cwe-m-cmoc---co-o—oooooooooooo= -0.13| 1.51} =-0.07 0.90 Grade 7 Monongalia County samples-===-===-=-=-----==-==c-=---------- -0.09| 0.83| -0.18 1.70 Control Samples--=--====m=mmmcc-occe-acmmm-ommmoo—o—o---- -0.63|'4.74| -0.12 1.11 Grade 8 Monongalia County samples-=--===---==c==--===cc=-o—o--oc-- -0.29| 2.42 -0.12 1.05 Control sampleS-----====e-=mmecccco—cocooooooomooooooocos -0.15| 1.17| +0.04 0.31 Grade 9 Monongalia County samples----=---=-=-======-=-=-c-----c=- -0.30| 2.10] -0.25 1.84 Control sampleS-----====se--mececccccocoomomooooooooomoo= +0.41| 2.48) +0.33 2.09 lgignificant at the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance of the difference be- tween means. the Arithmetic section. Here also, then, the Read- ing section might suffice as a brief estimate of school achievement. Relation to General Ability and Socioeconomic Status Table 46 gives the correlation of the WRA'T with the measure of general ability and with the socioeconomic status of the students’ parents. It is found again that for these samples one-third to one-fourth of the variance is common with the measure of general ability. Correlations with parents' occupational level, however, remain nonsignificant or quite low and in no instance account for more than 15 percent of the common variance. As for the elementary school sample, it can be concluded, therefore, that the WRAT is reasonably applicable to subjects of varying socioeconomic backgrounds. A similar conclusion with respect to levels of intelligence, however, must await further investigation. 35 Table 44. Grade level discrepancies between selected criterion measures from the Stan- ford Achievement Test and the Reading section of the WRAT, by subtest, grade, and sample for the junior high school samples Word Meaning Paragraph ; or Vocabulary Meaning Spelling Language Grade and sample D t E D E D t Total junior high Monongalia County samples-| +0.30 2,22 | 40.43 13.23 --- ---1+40.89 15.97 Control samples---=-ceeaa- +0.53 | 14.82 +1.32 10.91 | +1.39 | "13.37 +1.20 19.68 Grade 7 Monongalia County samples-| +0.32 1.66 +0.18 0.95 coo Jes +0.29 { 1.34 Control samples------==--- +1.29 | 6.45] +1.98 | '8.40 [+1.83| '10.00| +2.35 10.22 Grade 8 Monongalia County samples-| +0.14| 0.59 Control sampleS=--===ee--- -0.21 1.14 Grade 9 Monongalia County samples-| +0.41 1.57 Control samples-=-====ce-= +1.33 8.36 .31 1.36 | +0.25 1.05| +1.48 15.76 70 13.55 [40.43 | 12.74] +0.03 0.16 +0.76 | 2.99 | +0.51 2.06 +0.58 2.16 +1.32| 17.37 |+1.90| '12.10] +1.26 17.54 Isignificant at the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance of the difference be- tween means. IV. THE SENIOR HIGH SCHOOL STUDY BACKGROUND DATA The subjects for the senior high school study were obtained in Monongalia County by the ex- haustive testing of students in the upper three grades of a semirural high school, and quotas were completed by random sampling from the University High School. Similar quota sampling was used in one high school in each of the three control areas. Because of the demographic dis- tributions inthe sampling areas, the average occu- pational level of the parents was somewhat low in West Virginia and California. The remaining 36 two samples (Colorado and Wisconsin) were closer to, or slightly above, the national average. With respect to intellectual ability, rather close matching to national averages was obtained for the West Virginia and California samples, while the Wisconsin and Colorado samples showed above average ability levels, probably reflecting different patterns of high school dropout than was true for the principal sample. Table 47 gives the relevant data on parent's occupation, pupil ability level, days between individual and group tests, age at time of test, and grade level at time of test. It should be noted that for the Monongalia and control samples both grade and age levels have been matched to within less than one-tenth of a grade level. ADEQUACY OF GRADE LEVEL PLACEMENT Table 48 gives means and standard deviations for the WRAT raw scores, and similar data for the WRAT grade level scores are provided in table 49. All grade level estimates (with the ex- ception of the Wisconsin tenth grade sample) were below actual grade level placement. Regional discrepancies were again the most noteworthy. The West Virginia and California samples showed actual grade level placement underestimates ranging from two to three grade levels, while the remaining samples came very close to actual grade level. Magnitude of discrepancy from actual grade level and associated significance test re- sults are reported in table 50. Considering the above-average intellectual level for the Wisconsin and Colorado samples, it must again be concluded Table 45. lation of each Correlation of the Arithmetic and Reading sections section with the Arithmetic and Reading that the WRAT, in general, seriously under- estimates actual grade level for senior high school students. A fair consideration of the test author's position once again must include reanalysis of the data with respect to the notion that grade levels are not expected to show systematic increment because of different (and often automatic) pro- motion policies. The grade level equivalent at the 50th percentile corresponding to the average age of our grade sample was obtained from Jastak's manual, and discrepancies were recom- puted using these new levels as reference points. No significance tests are available for the re- vised discrepancies also reported in table 50. By inspection, however, it may now be seen that use of Jastak's conversion tables results in dis- crepancies which seem to reflect the intellectual levels of the several samples. Thus use of the conversion tables leads to obvious overestimates of grade level for the Wisconsin and Colorado samples. However, grade level estimates for the of the WRAT and corre- (A + R) composite score, by sample and grade for the junior high school samples (raw scores and grade level scores) a Arithmetic| Reading ' ith + ith Calle Reading wi A R| with A + R M Cc M C M C Raw scores Total junior high===wmemm=====-cccnnem——— .58 i | .78 .86 .94 +97 Grade 7----=====-------mmmmm-so----------o----= .60 44 .78 .69 +37 +25 Grade 8-------=-===--m---me-mm---o-o-----o-----o +38 .65 .70 .83 .84 «97 Grade 9---------m---memmeemmmmmmmoo-—-—m—m—mmm .58 .63 77 .86 .97 .94 Grade level scores Total junior high-e--sece-ececemcenneenan-—- .59 «70 +85 .89 +93 “25 Grade 7----=m=eemmmmeceeceeece-mo--o--coo---oooooo- +57 42 .82 .75 94 91 Grade 8---=-mecemmme—cecmeemmoc—ceossemmm—— es 45 .63 .81 .88 .89 .93 Grade 9---=-m-eeemmme--ecemcmmmmm-————ee-——————— .58 .64 .84 «92 +93 .89 NOTE: M-—=Monongalia County samples; C—control samples. 37 Table 46. occupational level of parent,by subtest, samples Correlation of WRAT grade level scores with general ability level and with sample,and grade for the junior high school Grade General ability level with: Arithmetic Reading | yTpfbmetic M c M| c| M c 49] .59| .56| .62 «39 .66 Grade 7==mm- oom moe eae +54 .51| .56| .63 +62 .68 Grade 8--=- moma .51 .60| .52| .60 .61 .66 Grade 9--===- mmm. .50 .35] .63] .38 .65 .40 Occupational level of parent with: Grade Arithmetic Arithmetic | Reading + Reading M C M C M C Total junior high==-==mc como __. +14 .20| .17| .32 .18 +29 Grade 7===m—m ooo oem. ————— +29 LOL 267 .12 +31 .09 Grade 8===m- momo. .18 .35( .30] .35 .28 .39 Grade 9===m--mm ome eee .08 .20| .08| .13 .09 +19 NOTE: M—Monongalia County samples; C—control samples, West Virginia and California samples now more closely approach their actual grade average. Perhaps some closure can be achieved by considering the discrepancies for the combined senior high school samples. Inspection of the totals in table 50 suggests that use of grade level estimates for senior high school students under- estimates actual level while use of Jastak's conversion table results in a slight overestimate, 38 PERFORMANCE ON THE MAT Data on the student's performances on the Metropolitan Achievement Test (MAT), the cri- terion variables for the senior high school sam- ple, were analyzed in twodifferent ways. Standard scores are available which indicate the student's performance as compared with the total high school sample, on which the test was standardized, Table 47. Means and standard deviations on selected background data, by grade and sample for the senior high school samples Days LL between A : 0 ge at Grade level Sloat Aniliey individual time of at time of Grade and sample parent and froup group test group test Mean | S.D. | Mean | S.D. | Mean S.D. Mean S.D. | Mean (S.D. Total senior high Monongalia County samples---=---=- 3.70 | 2.36] 4.14 | 1.05| 6.63| 10.51| 17.08 | 1.06] 11.85] 0.81 Control samples----------e-mcen- 4.83 12.57] 4.41] 0.94] -4.74| 3.22 16.92 | 0.85| 11.78] 0.51 Grade 10 Monongalia County samples--=-=---- 3.69 | 2.30| 4.17 | 1.08 6.79] 11.05| 16.09 | 0.62] 10.83 0.05 Control samples-=-----cc-eocmnu=n 5.11 | 2.64] 4.58 | 0.79] -7.18 1.02| 16.11 | 0.44] 10.90| 0.00 Grade 11 Monongalia County samples=--=-=---- 3.71 | 2.28] 4.14 | 1.10 7.02| 10.32| 17.03 | 0.80] 11.82 0.04 Control samples---=---cecommamn= 4.02 | 2.34) 3.94 | 0.87| -5.88| 0.81] 17.00 | 0.45| 11.90| 0.00 Grade 12 Monongalia County samples------=- 3.70 | 2.40 4.11 0.95| 6.11] 10.17] 18.04 | 0.67] 12.82| 0.04 Control sampleseeme---ccccccacax 5.55 | 2.46| 4.83] 0.98] 0.46| 0.53] 17.99 | 0.39} 12.90| 0.00 NOTE: S.D.—standard deviation. These scores are in the typical T-score form with a mean of 50 and a standard deviation of 10. On scores such as these, one would expect the eleventh grader to fall at about the average while the tenth grader should be below and the twelfth grader above the mean given for the total norma- tive population. A second type of score, the within grade stanine, permits comparison of the study's subsamples withnational norms. Means and stand- ard deviations in T-score form are reported in table 51 and their grade stanine equivalents are given in table 52. The Monongalia County sample fell at or above average on the subtests of Reading, Lan- guage, Language Study Skills, Social Studies In- formation, and Science Information, while it fell somewhat below the national average on Spelling, Mathematical Computation, and Mathematical Analyses. Underestimates of achievement of grade level for the Arithmetic part of the Wide Range 39 Achievement Test for these samples may there- fore be attributable to their generally lower level in mathematics achievement, As on the WRAT, the control samples from Colorado and Wisconsin exceeded the West Vir- ginia samples on practically all of the Metro- politan subtests, while the California sample per- formed slightly below the West Virginia sample. Care must be taken, however, to remember the higher average on the general ability index for the two control samples before interpreting these results. Comparison of the control samples to national averages requires consideration of the within grade stanines reported in table 52. The average stanine has a value of five. Hence, it follows that the Wisconsin sample was at about the national average on most subjects except Language, on which it was low, and Science and Social Studies, on which it was high. The California sample appeared to be at or near average on Reading, Social Studies, and Science and low on all other subjects, while the Colorado sample was at or about average on Spelling and Language and above average on all other items. The Monongalia County samples, finally, were all below average on Mathematics and Spelling, and the twelfth grade sample appeared low on all subjects except Science and Social Studies. RELATION OF THE WRAT TO THE MAT Table 53 gives the validity coefficients de- scribing the relation of the WRAT Arithmetic section grade level scores with the standard scores on the Metropolitan Achievement Test. Results for individual grade levels and combined high school samples are given separately for the Monongalia County and control groups. The boxed group of coefficients in this table represents the Table 48. Means and standard deviations on the WRAT, by subtest, grade, and sample for the senior high school samples (raw scores) Grade and sample Arithmetic Arithnetic Reading + Reading Number Mean S.D. | Mean S.D. Mean S.D, Total senior high Monongalia County sampleS==ee=mmeseeccccmceneoon Control sampleS-m=mmmmeecececcocmocecccccacaa Grade 10 Monongalia County sampleS--=s==-emmeeeeececcacona Control sampleS-eemeecccmcocmccccmccccccccce—a- Grade 11 Monongalia County samples=--==-memeeecccccacan- Control sampleS=ememeeeecemccccmccccccccccccccanaa Grade 12 Monongalia County sampleS-=e=ee-cemmeccecececann Control sampleSememeecccmacccmaacnas [PS 301) 28.30| 6.34] 55.21 | 13.04 83.52 | 17.38 269] 30.11| 6.75] 59.54 | 10.74] 89.62 | 15.50 95] 27.07| 5.93] 51.52 | 12.75| 78.59 16.24 102] 31.11| 5.92] 60.03 | 9.38] 91.14 13.21 103| 28.85| 6.22 | 57.43 | 13.40] 86.31 17.98 98] 26.36| 6.30] 55.98 | 11.67 82.34 16.06 103] 28.87| 6.66 | 56.40 | 12.18] 85.27 16.81 69] 33.96| 5.74] 63.87 | 9.45] 97.71 12.95 NOTE: S.D.-—standard deviation. 40 Table 49. Means and standard deviations on the WRAT, by subtest, rade, and sample for the senior high school samples (grade level A Arithmetic Arithmetic Reading + Reading Grade and sample Number Mean S.D.| Mean S.D. | Mean S.D. Total senior high Monongalia County samples--=----=----- 301| 9.48] 2.78] 10.29 | 3.16| 9.88 | 2.61 Control samples==-==----===--c-c--c-===- Grade 10 Monongalia County samples--=-=-=----=-=-- Control sampleS--r=wrem===-~rwcnaanm—- Grade 11 Monongalia County samples------------- Control samples--=ewmm==mmmmmawnnnnn.. Grade 12 Monongalia County samples--------=----- CONLYol SaMPLeS=— wm mmm =m vm me mmm orn ww oe 269 | 10.32] 2.99] 11.36 | 2.66 10.84 | 2.48 95| 8.90| 2.55] 9.40 |3.02| 9.13] 2.38 102 | 10.72) 2.71] 11.46 | 2.39] 11.09 | 2.19 103 | 9.72] 2.74] 10.87 | 3.24] 10.30 | 2.67 98 | 8.68] 2.65| 10.48 | 2.81 | 9.58 | 2.42 103 | 9.77] 2.93| 10.53 | 3.02 | 10.15 | 2.61 69 | 12.05 2.65] 12.45 [2.37] 12.25 | 2.08 NOTE: S.D.—standard deviation. correlations with the most pertinent criterion variables, the subtests involving subject knowl- edge of Mathematical Computation and Concepts and of Mathematical Analysis and Problem Solv- ing. The validity coefficients for these specific criteria for the individual grade samples range from .62 to .82 for the Monongalia County and from .66 to .77 for the control samples. Values for the combined high school sample are .68 and .77 and .73 and .78, respectively. These values are above correlations with the nonmathematics criteria and thus demonstrate construct validity for the WRAT Achievement section also in the high school sample. Validity coefficients describing the relation of the WRAT Reading grade level scores with the standard scores from the Metropolitan Achieve- ment Test are given in table 54. Here the most pertinent criterion variables would seem to be the subjects Reading, Spelling, and Language, although the topics Language Studies Skills and Social Studies Vocabulary are also clearly rele- vant, Correlations with the three most pertinent criteria range from .61 to .82 for the Monongalia County samples and from .49 to .82 for the con- trol samples. Again, construct validity seems present for the Reading section of the WRAT since correlations are generally higher for the lan- guage-related than for the nonlanguage-oriented subject matter criteria. Relationships were also evaluated between scores for the combined Wide Range Achievement Test and the standard scores on the MAT. Coeffi- cients describing these relationships are listed in table 55. As in the studies using the Stanford Achievement Test as the criterion, it is again found that use of the combined WRAT score im- proves prediction for the language-related cri- terion measures while it does not significantly affect the magnitude of prediction of the arith- metic measures. 41 Table 50. Discrepancies between observed WRAT grade level scores,actual grade level, and Jastak's age norms, by subtest, grade, and sample for the senior high school samples Discrepancy from actual grade level Joisorgpeney om Arithmetic Grade and sample | iui Reading + ipleimaeie Reading Arithmetic | Reading + Reading D t D t D t Total senior high Monongalia County . g samples--=-==------ »2.37 14.81 | -1.56 8.62 -1.97 13.13 -0.29 +0.52 +0.11 Control samples----| -1,47 18.12 | -0.43 2,6 -0.95 16.33 +0.57 +1.61 +1.09 Grade 10 Monongalia County y . samples--=-=-=------ -1.91 7.38 | -1.41| "4,56 | -1.68 '6.94 -0,65 -0,15 -0.42 Control samples----| -0.20 0.75 | +0.56 2.36 | +0.19 0.88 +1.12 +1.88 +1.51 Grade 11 Monongalia County J samples---====u--- -2.08 1.7.76 -0.93 12,92 | -1.50 5.42 -0.03 #+1.12 +0.55 Control samples----| -3,24 | "12,09 | -1.44| "4,91 -2.34 '9.59 -1.07 +0.73 -0.17 Grade 12 Monongalia County 1 . samples-====--=--=- -3.03 | '10.52| -2.27| ‘7.64 | -2.65| "10.27 -0.18 +0.58 +0.20 Control samples----| -0.85 12.66 | -0.45 1.58 | -0.65 2.60 +2.10 +2.50 +2,30 'Significant at or beyond the l-percent level of confidence. NOTE: D—algebraic difference; t—t-test of the significance of DISCREPANCIES BETWEEN WRAT AND MAT ESTIMATES While the correctness of grade level esti- mation may not be as crucial at the high school level as it is for children at the earlier ages, it is nevertheless important to assess to what extent, if any, the WRAT tends to overestimate or under - estimate school achievement as measured by an independent assessment procedure. No grade level estimates were available for the MAT, but it was possible touse Jastak's tables to convert grade levels into standard score form. Jastak's standard scores were converted into the conventional T-score form, and discrepancies were computed, even though the absence of ap- 42 the difference between means. propriate statistics in the manuals precluded formal significance tests. Tables 56 and 57 give the WRAT Arithmetic and Reading score means in T-score form and list the discrepancies in T-score points from the corresponding means on the Metropolitan Achievement Test. There is considerable varia- bility among grade levels and samples. The clearest picture merges when we consider the combined means for the total high school samples. Here it appears that the Arithmetic section on the WRAT slightly overestimated achievement in mathematics for the Monongalia County samples but was approximately close, on the average, for the control samples. The WRAT Reading section closely predicted average level for the Metro- politan Reading Test (except for marked under- estimate of the Colorado twelfth grade sample) but seemed to overestimate the Spelling and Language subtests of the Metropolitan by an average of one-half of a standard deviation. THE WRAT SECTIONS Interrelationship Table 58 gives the intercorrelations between the Arithmetic and Reading sections of the WRA'T for the high school samples. Substantial corre- lation exists between these sections for all sam- ples, and the finding for the elementary and junior high samples regarding the higher corre- lation of the Reading section with the combined Arithmetic and Reading score is replicated. In- spection of tables 55 through 57, however, sug- gests somewhat better definition in criterion prediction for the two forms of the test. Thus, while the Reading section presents a better choice if the battery has to be shortened, there appears to be greater justification for retaining both subtests of the WRAT at the high school level than at the lower grades. Relation to General Ability and Socioeconomic Status The correlation of the WRAT parts and com- bined score with the measure of intellectual ability and occupation of parent is given in table 59. For the high school samples, correlations of the WRAT with occupational level, as the estimate of socioeconomic status, are nonsignificant, or trivial, and present evidence of the utility of the test for children from varying socioeconomic backgrounds. Correlation with intellectual ability is again quite substantial, although some decre- ment in the relationship occurs at the twelfth grade level. V. VALIDITY OF THE WRAT AT EXTREME ABILITY LEVELS WRAT PERFORMANCE AT EXTREME LEVELS In the previous section substantial corre- lations were reported between the WRAT scores and the measure of general ability. These findings raised serious questions as to the suitability of the WRAT at extreme levels of intellectual ability. In order to handle this problem, special studies were conducted on samples of subjects at both the lowest and highest levels of the ability range. For this study subjects were pooled from the Monongalia County and the control samples and subsamples were pooled wherever comparable forms of the criterion tests were available. As a result, data are presented on four samples. Two of these represent students from the fifth through the ninth grades, all of whom took the Stanford Achievement Test. The other two sam- ples represent students from grades 10-12, on whom we had comparable data on the Metropolitan Achievement Test. In each set, one sample rep- resents students with IQ's of 80 or below, while the other sample represents students with IQ's of 121 or above. Table 60 shows the mean age and grade level placement for each of these samples, as well as the WRAT grade level scores. As would be ex- pected, mean ages are higher for the low-ability groups and the average grade level scores on the WRAT are much below actual grade level for the low-ability groups. The two high-ability groups are substantially above their expected grade level, but of course, the discrepancy here is not as great for the low-ability group. In the analysis of the validity of the WRAT for the extreme groups, the matter of the relation of the WRAT to the most pertinent criteria from the Stanford and Metropolitan Achievement Tests will again be attended to and, in the case of the SAT, the discrepancy between grade levels will be considered. 43 Table 51. Means and standard deviations on the Metro subtest, and sample f politan Achievement Test,b or the senior high school samples (standard scores J grade, Total senior high Grade 10 Grade 11 Grade 12 Subtest and sample Mean 8.0. Mean S.D. Mean S.D. Mean S.D. Reading Monongalia County samples---| 52.70 | 13.76 50.54 | 13.18 | 54.83 | 13.69 | 52.56 14.02 Control samples----=--caaaa- 56.41 | 13.86 52.39 [12.80] 52.61 | 12.60] 67.74 10.41 Spelling Monongalia County samples---| 47.76 | 16.16 43.96 (14.79 | 48.59 | 16.17 | 50.45 16.71 Control samples-------ccme-- 50.96 | 15.84 || 49.72 | 14.57 | 47.49 | 16.71] 57.71 14.24 Language Monongalia County samples---| 51.06 | 16.21 || 48.63 | 15.63 | 51.87 | 15.67 | 52.49 16.99 Control samples-----=-eceaa- 52.15 | 14.12 || 46.76 [12.52 50.24 | 12.95] 62.83 12.08 Language Study Skills Monongalia County samples--- -——— -—— -—— ---| 51.18 | 18.27] 51.30 18.98 Control samples------===eea- 56.84 | 15.56 53.16 | 14.23 | 52.15 | 14.45] 68.93 12.30 Social Studies Skills Monongalia County samples--- -—— -— -—— -—— -—— —— —— -——— Control samples-------ececeua- -—— -—— 53.16 | 14.23 -—— ---]167.68 13.77 Social Studies Vocabulary Monongalia County samples--- -— -—— -—— ——— ———- am ew oe Control samples-------ecce-a- 58.48 | 17.00 || 54.34 | 15.37 | 54.07 | 15.56 | 70.87 | 15.07 Social Studies Information Monongalia County samples---| 52.66 | 15.09 || 49.66 | 13.67 | 53.61 | 13.94 | 54.47 16.91 Control samples----=-===u-ca- -—— -—- 54.12 | 12.33] 56.08 | 13.64 -——— -—— Mathematical Computation and Concepts Monongalia County samples---| 46.54 | 15.54 || 43.39 [13.62 48.87 | 14.58 47.12 | 17.52 Control samples-=--=---cceeeo 52.83 | 16.47 51.24 | 13.47 | 46.41 | 15.77 | 64.29 15.55 Mathematical Analysis and Problem ss Monongalia County samples---| 46.19 | 17.51 44.40 116.94 | 46.37 | 16.60 | 47.66 18.72 Control samples-------ceceaa- 53.59 | 17.55 || 51.15 | 14.65] 46.99 | 16.26 | 66.59 16.38 Science Information Monongalia County samples---| 51.62 | 14.82 52.74 | 14.85] 51.94 | 12.38 | 50.26 16.79 Control samples-------ccea-- -—— -—- 56.77 | 13.62 | 51.77 | 15.97 -— -—— Science Concepts Monongalia County samples--- -——— -— -—— -—— -—— -—— - -— Control samples------=cecea- -——— -—— 57.87 | 14.33 -—— ---167.29 15.34 NOTE: S.D.—standard deviation. 44 Table 52. Means and standard deviations on the Metropolitan Achievement Test,by grade, subtest, and sample for the senior high school samples (within grade stanines) Total senior high Grade 10 Grade 11 Grade 12 Subtest Mean | S.D. Mean | S.D. | Mean | S.D. | Mean | S.D. Reading Monongalia County samples---=-=-=-=-- 4.73 1.87 || 4.92] 1.77 14.98 | 1.94] 4.31 1.82 Control samples-=---cecceecec-cce==- 5,33 1.80 5.15] 1.70 | 4.73 | 1.76 | 6.43 1.47 Spelling Monongalia County samples=-----=-==-- 4.21 2.07] 4.12] 2.02]4.32| 2.12] 4.17| 2.05 Control samples=e---==ceeeccccccca= 4.65 2.08 4.87] 2.034.141] 2.13 5.04 1.94 Language Monongalia County samples----=--===-- 4.47 2.03 4.65] 1.99 | 4.58 | 2.02 | 4.19 2.04 Control samples-e-e=-eeececccccan== 4.58 1.725 4.25) 1.64 | 4.30 | 1.76 | 5.48 1.59 Language Study Skills Monongalia County samples---===-=--- -—- --- e-=| ---|4.48| 2.31] 4.15 2.31 Control samples=====-csceeeccccce== 5,23 1.96 5.15| 1.85 4.57 | 1.94] 6.30| 1.68 Social Studies Skills Monongalia County samples--=-==----- —— -—— -—— -—— -—— -—— —— _—— Control samples-===-=-=---==-c-cc=--- -—- -—- 5.30 | 1.72 -—- ---] 6.42 1.73 Social Studies Vocabulary Monongalia County samples--======-=- —_—— -—— -— -—— -—— -—— -—— -—— Control samples---=====-===---==c==- 5.80 2.15 6.01 2.07 | 4.93 | 2.04] 6.71 1.94 Social Studies Information Monongalia County samples--=====--- 4.76 | 1.92| 5.17| 1.77 | 4.73 | 1.72] 4.41 2.15 Control samples==-=-c===cce-ccm==== -— -— 5.76 | 1.57 | 5.12 | 1.83 -—— -—— Mathematical Computation and Concepts Monongalia County samples---=--=-==-- 4.23 1.881 4.07| 1.77 ]4.45| 1.77] 4.16 2.06 Control sampleS-====e-ee=-ceccce==- 5.01 2.06 5.12 1.90 | 4.13 | 1.93] 6.09 1.90 Mathematical Analysis and Problem Solving Monongalia County samples---=-==-=---= 4.241 1.99 4.27) 1.97 |4.13|1.91]4.33| 2.07 Control samples-ee=ceececccccnea===- 5.13 2.02 5.18] 1.80] 4.27 | 1.90] 6.29 1.87 Science Information Monongalia County samples---===-==-- 4,84 1.91 5.34 | 1.94 | 4.70 | 1.68 | 4.53 2.01 Control samples--=ce==-ecccccnce==- ——— -==|l 5.85] 1.80]|4.73| 2.09 =--- —— Science Concepts Monongalia County samples----====--- -—— _— -— -———] --- wo i soi Control samples=-==-=ce-e--cceee==—= -— -——- 5.52| 1.75 -—— ---16.01 1.86 NOTE: S.D.-—standard deviation. 45 Table 53. Validity coefficients describing the relation of the WRAT Arithmetic grade level scores with the standard scores on the grade, sample, and subtest for the senior high sc Metropolitan Achievement Test, by hool samples Total senior Grade 10| Grade 11| Grade 12 Subtest high M Cc M Cc M Cc M Cc Reading=====- moomoo mee. «B21 «57 .58 | .50| .58| .55| .68 .50 Spelling==== moomoo. .58 | .54 51) .56| .50| .50| .66 43 Language-======-c=cocmm mmm. .66 | .58 64 | 46] 57 69) .74 «B) Language Study Skillg==-=-cececaooo oo ___- ml 550 ===] .45] .52| .57| .69 .62 Social Studies Skills-==-=cocoomemooao___. ——— --- —== | 46] ===] ===] --- .56 Social Studies Vocabulary-----=-eeocoaoooo__ -— .59 === .62| ===] .60| --- +52 Social Studies Information--------ococoo_o_. 59 --- .54 | 43] .53| .56 | .65 -——— Mathematical Computation and ConceptS------- 77] .78 70 L771) .76] .76 | .82 «73 Mathematical Analysis and Problem Solving--- | .68 | .73 | .74 | .66| .62| 69 .69 74 Science Information---=-ceceemmooeo oo oL__. S52 --- 51 | .52) 51 .49) .58 | --- Science Concepts===--====ccmmmmooooo_. ——— === 1.50 === | ===] --- «37 NOTE: M—Monongalia County samples; C—control samples. Table 54. Validity coefficients describing the relation of the WRAT Reading grade level scores with the standard scores on the Metropolitan Achievement Test, by grade, sample, and subtest for the senior high school samples Total senior Grade 10 | Grade 11| Grade 12 Subtest high M Cc M Cc M C M Cc Reading-=====wcomm meee. 61] .61 .66 | .57] .61| .65] .56 .49 Spelling--=== mmm come 131 «76 .69 | .69| .82| .82] .66 .69 Language--===== ecm... .65| .65 72 | 67] .62| .74] 63 | .49 Language Study Skills-==we=-cocmaooooaa __. ===] .61 || ---1| .55] .62| .67] .53| .48 Social Studies Skillg====-cmeceooaaa oo ____ ———] --- === 40] === | === | --- .13 Social Studies Vocabulary=---=-=--cc-aeo_oo_-- ===] 62 ---| .54 |---| 67] --- .58 Social Studies Information==----==-cocoaao-- .56 | --- 62 431.59 | .52} 48 | --- Mathematical Computation and Concepts------- .53| .54 .56 | .51| .57 | .58 | .46 .30 Mathematical Analysis and Problem Solving---| .49| .53 | .51 | .46 C48 | 44) 48 | (31 Science Information=--=--=eceooaoao 2 _. 56 | --- .63] .56 | .63| .44| 51 | --- Science Concepts=====meeeaooommmo oa. -———] = === .54 | === | ===] .48 -——— NOTE: M—Monongalia County samples; C—control samples. 46 Table 55. Validity coefficients describing the relation of the combined WRAT Arith- metic and Reading grade level scores with the standard scores on the Metropolitan Achievement Test, by grade, sample, and subtest for the senior high school samples Total senior Grade 10| Grade 11 | Grade 12 Subtest high M Cc M Cc M Cc M Cc Reading-------=-=--m--cmemccmommmme mmm mm m= .70 | .67 .73 | .63| .67| .68] .70 .60 Spelling--=-=----=-=-meee=mee= meme —————————— +15] 73 L720 72 75 75) .75 .67 Language-=--===-==-=-=-==-=c==-=---------=----= 751.70 .80 | .65| .67| .80| .78 .67 Language Study Skills----------------=c--o-- --=-1 .68 ---| .58| .65| .70 | .69 +67 Social Studies Skillg===--==c---=ccmcmo-nu== ——= | =-- -== 1.50) === | === | === 44 Social Studies Vocabulary---------------=---- ---| .69 | ---| .61] ===] .72| --- .66 Social Studies Information--------=---------- .66 | --- .68 | .50| .63| .60| .64 | --- Mathematical Computation and Concepts------- 234 .73 22 2781 474 | +75) «73 .64 Mathematical Analysis and Problem Solving---| .66 | .69 .72| .66| .60| .63 | .67 .65 Science Information-=-=---==-=c-c-c--c-o-o--= 62 | --- .68 | .63| .65| .52] .62 --- Science ConceptS-==--=--=-==-=c==-c-=c---=-=-= ———| === --- | 60} === | === | --- .64 NOTE: M—Monongalia County samples; C—control samples. Table 56. T-score discrepancies between selected criterion measures on the Metropolitan Achievement Test and the Arithmetic section of the WRAT,by subtest, grade,and sample for the senior high school samples WRAT Mathematical | Mathematical Arichmeric Computation Analysis and Grade and sample ecore in and Concepts | Problem Solving T-score form D D Total senior high Monongalia County samples--------====----- 49.3 +2.8 +3.1 Control samples-=-=---==-==c-ee-cccmccaao—- 53.0 +0.2 -0.6 Grade 10 Monongalia County samples--------=-=------ 47.3 +3.9 +2.9 Control samples-=-==--==-ee-coco-cccacoo-a 53.3 +2.1 +1.1 Grade 11 Monongalia County samples---==----=-=------ 50.0 +1.1 +3.6 Control sampleg==-=-----=-=cemcommcocceannaa 46.0 -0.4 -1.0 Grade 12 Monongalia County samples---=---------c---- 49.3 +2+2 +1.6 Control samples===----===-ee-occcccmcnanoa 57.0 -7.3 -7.6 NOTE: D—algebraic difference. 47 Table 57. T-score discrepancies between selected criterion mea Achievement Test and the Reading section of the WRAT, sures onthe Metropolitan by subtest, grade, and sample for the senior high school samples .. Reading | Spelling | Language Grade and sample score in Tegoove D D D orm Total senior high Monongalia County samples-----=--coccmmeao__ 52.0 -0., 7 +4.2 +0.9 Control samples===-----ccmmmmmm eee 56.0 -0.4 +6.0 +3.8 Grade 10 Monongalia County samples==-=-----ceeomaoo__ 48.7 -1.8 +4.7 +0.1 Control samples===-==-ccccmmmm meee 56.0 +3.6 +6. +9.2 Grade 11 Monongalia County samples-====-==ceceamoooao 52.0 -2.8 +3.4 +0. 1 Control samples=-==--=-ccommmmmm meee eeemao 52.7 +0.1 +5.2 +1.5 Grade 12 Monongalia County samples--=---=--=--c-cc---- 52.0 -0.6 -1.5 -0.5 Control samplesS====--cmc- cmc mecca 58.3 -9.4 -1.6 -4.5 NOTE: D—algebraic difference. Table sample and grade for the senior high school samples (raw sc 58. Correlation of the Arithmetic and Reading sections lation of each section with the Arithmetic and Reading of the WRAT and corre- (A + R) composite score, b ores and grade level scores hriihaetic Arithmetic Reading Resin with A + R| with A + R Grade g M Cc M Cc M Cc Raw scores Total senior high--==--ccmommmo eee 1 .54 .78 .81 .95 +93 Grade 10-===-- === moomoo. Jab 46 o 2d .78 .94 292 Grade ll-===cm-m momo eos .61 .56 .81 .80 .96 +95 Grade 12-=m= =m cmm ooo +35 .38 .80 «73 .94 .90 Grade level scores Total senior high====-ccccmmmmmmaee oo .54 .54 .86 .89 .89 .86 Grade 10-==- === mmm mmm - 45 47 .82 .88 .88 .84 Grade ll-==== comm moe o +39 +57 .87 .88 +91 .89 Grade 12-===- momo oe eee .54 .36 .87 +35 .88 .80 NOTE: M—Monongalia County samples; C—control samples. 48 Table 59. Correlation of WRAT grade level scores with general ability level and with occupational level of parent, by subtest, sample, and grade for the senior high school samples General ability level with: : Arithmetic Grade Arithmetic | Reading | Reading M C M C M C Total senior high-----=-cce-c-ccmcmmmomm-oono .58| .61| .64 | .61 .70 .70 Grade 10===e-c-==m=-e-e-em-mm---e---sooomoooosssoss .63 .56| .70 | .66 .78 yo Grade lle---cee-memmcem-ceemmmmeo--e-sooooooososSoosos .60 .581.73] .66 +15 .70 Grade 12-=-e-e-cmemmc-ee-semmceo-s---osooooossssoos .56 45) 54 | 35 .62 .49 Occupational level of parent with: Grade Arithmetic Arithmetic | Reading + Reading M Cc M Cc M Cc Total senior highe==-c-c-eccccccccemccnconc=- «13 .23]1.09] .11 12 .20 Grade 10=--=-s=-m-e-m-es-eece---e-s---moooosssoSSos .05 L141 .10] .18 .10 .18 Grade lle----=-e-eesceecece-e--mcossss-eoeooososooSoes .10 L041 .09 | .07 +10 .02 Grade 12----==ee=mcememcc-mce-eooe-s-c-soeoossssosoos «21 .281.09 | .02 .17 .19 NOTE: M~—Monongalia County samples; RELATION BETWEEN THE WRAT AND THE CRITERION VARIABLES Table 61 gives validity coefficients for the group of extreme intellectual ability. It may be seen that at these levels the WRAT works reason- ably well in predicting performance on the Stanford Achievement Test at both high and low ability levels and that further evidence of construct validity is presented by the fact that the pertinent criterion variables correlate higher with the appropriate section of the WRAT. C—control samples. Less favorable results occur in the relation- ship of the WRAT to the Metropolitan Achieve- ment Test for the groups drawn from the senior high school samples. While there is significant prediction in the expected direction for the high- ability samples, validities are not as high as one would hope for. Of more concern is the fact that the Reading section of the WRAT completely fails to predict relevantcriterion variables for the low- ability group. It is true that this group is rep- resented only by a small sample (N=25). How- ever, modest correlations are yielded by this 49 Table 60. Means and standard deviations on the WRAT, time of test for groups of ext age at time of test, reme ability and grade level at Age at time | Grade level | Arithmetic a A+R of test at test grade level de 1 & 1 | grade level Ability group Number grade Jove Mean S.D. | Mean S.D. | Mean S.D. | Mean S.D. | Mean S.D, Low ability, grades 5-9--=-ooooo__- 521 13.76| 1.62 1.7% | 1.38 5.25 | 1.36 5.04 | 1.82 5.14 1.35 High ability, grades 5-9---=-c-oooo- 51) L2,17] 1.16 7.26 1.19 9.07 | 1.88] 10.74 2.57 9.90 2.00 Low ability, grades 10-12---------- 251 18.04 1.46] 11.78 | 0.67| 5.92 | 1.46 5.36 | 1.79 5.64 | 1.26 High ability, grades 10-12---------- 5916.87 | 1.02| 11.93] 0.87 12.92 | 0.64] 13.67 | 1.96 | 13.30 1.79 NOTE: S.D.--standard deviation; A + R—Arithmetic and Reading. Table 61. Correlation between the WRAT and selected criterion measures, by subtest for groups of extreme ability Arithmetic Arithmetic Reading + Subtests of the Stanford Reading and Metropolitan Achievement Tests Low | High | Low | High | Low | High Stanford Achievement Test Paragraph Meaning=-==---c comme o .34 .38 .61 .50 «39 .50 Word Meaning===-== === cco mmm eee .10 +43 .60 .64 .45 .62 Language -==== === c= cme - .16 +37 .50 74 42 74 Arithmetic Reasoning and ConceptsS=-==---=-cmoeeo- .58 .61 .38 v37 +35 .65 Arithmetic Computation--------ccooommmoo_o .60 +62 .24 .60 46 .68 Metropolitan Achievement Test Reading=-====- comm mmm +17 23 | [~.17 41 -.02 40 Spelling====-m =m mmo o .23 val .00 .64 .13 «32 Language--===-=c == == ccm eee 27 41 .14 .43 +26 .54 Mathematical Computation==---=--ccoomaeooo ooo L450 .72|1 .06 | .26] .30 .67 Mathematical Analysis and Problem Solving-------- «35 .64 .06 «28 +25 .63 50 sample between the Arithmetic section of the WRAT and appropriate criterion variables. The use of the Reading section of the WRAT at the high school level for students of low ability must therefore be viewed with great caution. DISCREPANCIES BETWEEN WRAT AND SAT GRADE LEVEL SCORES A final analysis of the performance of the extreme ability groups concerns the discrepancy of grade levels as estimated by the WRAT from the grade level estimates provided by the group achievement test battery. Table 62 presents data on the most pertinent criterion measures. There seems to be a tendency for the Arithmetic section to underestimate the mathematics cri- teria. However, the magnitude of the under- estimation is more serious for Arithmetic Com- putation than Arithmetic Reasoning and is probably significant only for the high-ability group. The WRAT Reading section for the high- ability grouptends to overestimate the language- related skills by more than one grade level. For the low-ability group, however, WRAT estimates are quite close to the criteria for Paragraph and Word Meaning, but again the WRAT overestimates performance on the SAT language subtest. These Table 62. Discrepancies between the WRAT grade level scores and the Stanford Achievement Test grade level scores, by subtest for groups of extreme ability Low ability | High ability group group SAT subtest Mean D Mean D WRAT Arithmetic test Arithmetic Reasoningand Concepts--- | 5.33] -0.08 8.68 -0.39 Arithmetic Computation- | 5.521 -0.27 1 8.23 -0.74 WRAT Reading test Paragraph Meaning----- 4,70 | 40.34 | 9.34 | +1.40 Word Meaning- | 5.19 | -0.15| 8.84 | +1.90 Language----- 3.74 | +1.16 | 9.32 | +1.42 NOTE: D—algebraic difference. findings are, of course, quite similar to those re- ported for the total sample, and they suggest that the WRAT canbe considered applicable to extreme ability levels for the elementary and junior high school children. VI. CONCLUSIONS The basic questions raised in this study in- volved the validity of the WRAT as a brief measure of school achievement and its adequacy for accu- rately predicting actual school performance as measured by conventional, comprehensive achievement measures. On the first issue it seems fair to conclude that the Arithmetic and Reading sections for both Levels I and II of the 1963 Revised Wide Range Achievement Test have reasonably good construct validity as judged by their relation to conventional group school achievement tests. While there is a considerable range inthe magni- tude of validity coefficients depending on the level and geographical region involved, there is sufficient evidence of substantial correlation with criterion measures at every age level investi- gated to consider the WRAT a satisfactory brief estimate of school achievement. Adequacy of the WRAT has alsobeen investi- gated at extreme levels of ability, and it is con- cluded that the WRAT is quite satisfactory with high-ability students. With respect to students of low ability, the WRAT still seems satisfactory except for the use of the Reading section with high school students, where validity seems in doubt. Since the correlation of the WRAT with level of parental occupation is quite low, it may further be concluded that the test is applicable for children with widely differing socioeconomic backgrounds. The question of grade level placement is rather complex. Here the WRAT must be rated as varying from being satisfactory to being in 51 considerable error, depending on the criterion used and the level at which the test is applied. Level I of both the Reading and Arithmetic sections of the WRAT overestimates bothactual grade level and achievement on criterion measures from the Stanford Achievement Test. The tendency for the WRAT to overestimate is particularly serious for the Reading section, and since it is found in various area samples it cannot be dismissed as being due to geographic peculiarities. Level II of the WRAT, on the other hand, tends to underestimate actual grade level but is quite close in predicting achievement levels on the Stanford Achievement and Metropolitan mathe- matics-related subtests. The WRAT Level II Reading test overestimated actual grade level for the junior high students but underestimated it for the senior high students. Likewise, per- formance on the SAT was underestimated, while performance on the Metropolitan criterion vari- ables was overestimated, Consideration of Jastak's age norms helps in some instances, such as providing more accurate grade level assignment at the junior high school level, but in other instances use of his tables increases the reported discrepancies. Analysis of the interrelation between the WRAT Arithmetic and Reading sections and their relation to criterion variables suggests that it would be possible to rely upon the Reading section as the sole achievement estimate at the elementary and junior high levels. However, such reduction of the achievement estimate would seriously re- duce the accuracy of the estimate for mathematics- related skills, particularly for the senior high school levels. Replication of our analyses for the Monon- galia County samples with the geographically dispersed control samples produced additional evidence of test validity but confirmed further that the grade level placement provided in the test manual must be used with caution, and may be subject to considerable geographical variation due to different educational policies and ability distributions. In summary, it may be suggested that the Arithmetic and Reading sections of the WRAT provide useful estimates of school achievement but that restandardization of raw scores and their grade placement equivalents on the basis of the Health Examination Survey data would be desir- able. Such restandardization ought to provide separate norms for broad geographic areas and, to be most useful, should provide separate norms for various age levels. 000 52 APPENDIX | LEVEL | OF THE WIDE RANGE ACHIEVEMENT TEST FORM USED IN THE ELEMENTARY SCHOOL STUDY Personal Data The four lines at the top of the title page should be carefully filled out before the test is begun. This section provides necessary statistical information. The following uniform procedures should be observed in completing the blanks. Name: Print last name first, then first name and initials. Never assume that you know how to spell a name (not even Smith). Have S spell or write it for you on the line provided on the test form. A correct name may save much time in filing and finding records when needed. Birthdate: Example: 10-18-1955 for October 18,1955. M. F.: Encircle M for male; F for female, Chronological Age: List completed years and months up to age 15 years, 11 months. For example, a child bornon7-21-1957 was 6 yrs.,3 mos. old on 11-15-1963. At 16 yrs. and above, list age in years only, using the year completed on the last birthday. A person born on 10-18-1943 was 19 yrs. old on 6-5-1963. School: Write down name of school attended at the time of the test. Grade: Enter the grade he is attending at the time of the examination in the case of school children. Date: Always record the date on the test. Example: 10-15-1960 for October 15, 1960. Examiner: Print name of person administering test. DO NOT COMPLETE ANY OTHER ITEMS. Now have the child write his name on the line below the little boxes on the first page. Test Instructions, Level | Reading E should acquaint himself with the pronunciation of the words in the list. The pronunciation guide is provided for the examiner on page 6. The transcription symbols are those found in Webster's New Collegiate Dictionary. The E may use other standard dictionaries or the symbols of the International Phonetic Association in learning to pronounce the words. Since this is primarily a reading test and not a test of speech or diction, unusual pronunciations due to colloquialism, foreign accent, and defective articu- lation are accepted as correct. An incorrect answer is any misreading due to improper sequence of letter sounds, confusion of phonetic values, and misplaced accent. Always begin the administration with the word pronunciation test (75 words). Two copies of the test form may be used, one for S to read from, and one (with personal data filled out) for E to record on. Point to the first word ''cat' and say: Look at each word carefully ana say it aloud. Begin here (point) ana vead the words across the page so I can hear you. When you finish the first line, go to the next line and then the next. In the case of young children (5 to 7 yrs.), each word should be pointed to with apencil while S attempts to read. Time: 10 seconds per word. The reading part should be administered with as few interruptions as possible. Any clearcut response should be accepted and scored as either right or wrong. The first time an error is made, S is asked to say the word again. His response is scored right, if he corrects himself on the second trial. From then on, the first response is scored as either right or wrong, unless S spontaneously corrects the error he has made. If the response is not clear, E may ask S to repeat the word. The E should not intimate, by either motion or emotion, that he is dissatisfied with the answers. Spontaneous corrections are credited, but teaching, coaching, or questioning should be avoided. The reading speed may be controlled by E. Saying 'mext'' at the end of the time limit of 10 sec. is one way of controlling the rate of performance. Refusals to read within time limits should not always be accepted as evidence of failure. If S hesitates or says "I don't know this word'', E should encourage S to try the word anyway or ''take a guess'' at it. 53 Testing Limits: 12 consecutive failures. Recording: a. Underline the first letter if the word is cor- rectly pronounced. Example: cat, block. . Cross out the first letter of the mispronounced word. Example: gat, Block. . If S first mispronounces the word, then corrects his error, cross out the first letter and under- line the second letter of the word. Example: gat, Block. Score right. . If S first pronounces the word correctly, then mispronounces it, underline the first letter and cross out the second letter of the word. Example: cdt, block. Score wrong. On the reading test, some Ss tend to skim over the words or produce a response that sounds superficially correct. The E should be alert to these near successes and score them wrong, or ask S to repeat if no clear- cut decision can be made. Examples of such near successes are: bik for blk, &kskap for &skap, humiditi for 5h. CAVY / v wv / wu. /, Vo humiditi, kontempori for kontemporari, akov'for al'kov, _ oe 4 —. v kSntémplius for kontémptiits, beni for bénin’ protiberins for protubérins, séntri’figal for séntriffigal, abiemil for ibizmal, susingkt for suksingkt’ etc. Pre-Reading Section Whenever failures occur in the first line of the reading test, the three pre-reading parts of the sub- test are administered as follows: 54 1. Naming 13 letters: Say, Reaa these letters aloua. What is this or What ao you call this? (Point to the first letter in the second row of capital letters printed above the word list: A B O, etc.) Point to each letter consecutively as S reads them. Time: 10 seconds per letter . Recognizing 10 letters: Cover the word list with a sheet of paper, point to the first letter (A) in the top row on page 4 of the test form and say: Fina one just like this down here (pointing to the row of letters underneath). The instruc- tions may be repeated if necessary. Each letter should be pointed to. Time: 10 seconds per letter . Two letters in name: Point to the first letter in the name which S has written on page one of the test form and say: What do you call this letter? or What is this letter? Ana this one? (pointing to the next letter). The E may stop after the second correctly named letter. Time: 10 seconds per letter Recording: Underline letter correctly named or rec- ognized, cross out letter incorrectly identified or named within time limits. Ww nN S22 ABBR BB BR LOW DW WWW WW AUR WN OOO IU DWN — NMBA MN = bt me bm ee 0 00 Ad Oh 1h OO BO NONE WN OOO NANE WD —O + . «oe os eo. nN oo oe Level I—Pronuciation Guide for Reading Test Caluvrirrnnneereennann, kat LO sunumvie ssvanvisansiinne too SEE ...iiieiriiiinnninann, s€ BOOK, vs sssmvisininsansnns bdBk big...oiviiiiiniiinn, big CAL yu veins navi dimimmmee et WaS....oeuvunernannnnnn. woz Bed ino nnnisiviiii sian nes réd him.............ceel, him letter. .................. 1ét’er OPeN......cvvvnrrennnnnn. 5’pén BOW...iviessennenevorase hou then.............c...... thén EOP. cnr srrrrnmans dép WOrK ....oovvnvninninnn, wiirk JAY vivir cniniesmnennae jar awake................... a wak’ BIZE soncniiinvmidinensare siz 15) 01) ER spel HPuss sicrsnrsinizasssrans lip block......cccu.u...... bi8k weather ................ weéth’er EVEN... ievueeviunns g'vén 11:0 fing/ger should .................. shood CH suniiimsnmennrenress kiif felt. ovvvveeerirnnnnnnn. felt Sralk i. iunverirenensreer stok WBY cs mmmmummermiminraia tra huge........ccvvvenn... huj approve................ a proov’ lame.................... lam PLO sisvisninsssornnesann pide SUCK . uvnn manmnvins struk quality.................. kwdld ti BOUL ons mmmmwmwnsnsinhs sour UIge...civivniiinnnnnnnn, Urj abuse................... a biz’ or a bis” . collapse................ k8 18ps” . exhaust................. gg 20st” bulk ....covuiiinnen blilk . esidence....eernia réz/i déns . clarify.................. klar fi . humidity ............... hit mid’Y of imply ......cocvvuennnn.. im pli’ . Quarantine,............ kwdr/n tén . threshold.............. thrésh’sld or thrésh’hold ~ 48. ZIULLON. .. i. vaviianvennss gltit”’n 49, Tecession......crcvrees ré sésh’ un 50, participate.....ccss pir st 1 pat 51. horiZon.....cccesvesss hd riz 'n 52. emphasis .............. ém’fd sis AL NY 53. AeronautiC,......sceses a er 6 nd tik or Ar & no tik 54. intrigue................ in reg 85. luxurious .:ccssesvsss liiks =u £5 us or lugz- 56. endeavor............... én déy’e: er 57. PELSEVErE ......cieies pur ‘se ver’ 58. rescinded. ............ re stnd’&d_ 59. discretionary......... dis krésh’ un er i 60. Mitosis ....cuus ever. mi 1 osls or mi- 61. repugnant.............. re ple 'nént 62. putative... cous wavennns pu’ta dv Y 63. rudimentary .......... 156 di mén "ta ri 64. NETeSY... cove: csovains hér of 65. USUTP..ecvevenirnnennnns i zirp’ or u sturp’ 86. NOVICE civirvervensinsons nov 18 67. audacious.............. 6 da’shiis, 68. anomaly... sverenessnsy a nom’ ‘a i 69. seismograph.......... s1z/md raf 70. idiosyncrasy.......... la} 8 3 sing ’kra si 71. itinerary............... T dn'er 4&4 or Tener ard 72. Spurious .....civesnsres spl YY us 73. miscreant............. mis it 74. aborigines............. 8b’ 51 nez 75. pseudonym............ ST nim or psu’dd nim Level |—Arithmetic This test is composed of an oral and a written part. The oral part of the subtest consists of: 1. Counting 15 dots 2. Reading 5 digits 3. Showing 3 and 8 fingers........ccviuiiniininininnininnnee. 4. Telling which number is more: 9 or 6; 42 or 28 ....... 2 = 5. Three oral addition and subtraction problems ................. 3 20 points The written part consists of 43 computation prob- lems. Test Instructions: Begin the testing with the written computations. In examining young children (5 to 7 yrs.) point to the first problem ( 1 + 1 =) and say: Read this. If the problem, including the signs, is read correctly, ask: "What is the answer?' When the answer is given, say: Write it down on this line. Then say: Now vead this (pointing to 4 ~- l=) ana put the answer on the line (point). Next read this (pointing to 6 + 2) andputthe answer undev the line. Then read all the other prob- lems in this row (pointing) ana write your answers on or under the lines, If the child is unable to read the first problem (1 + 1 =), discontinue the written part and administer the oral parts according to the instructions outlined below. Children of ages 5 to 7 yrs. and persons who obtain a score of less than 7 points on the written part, are given the oral parts of the subtest. 1. Counting 15 dots: Point to the dots printed at the top of page 2 of the test form and say: Point with your finger ana count these dots one by one beginning herve (S's left) ana going this way (moving to the right, motion). Count them aloua so I can hear you ana tell me how many aots there are. 2. Reading Numbers 3, 5, 6, 17, 41: Point to the numbers (printed upside down on the form) and say: Read these numbers. What is this? (point- ing to the 3). Ana this. Etc. 3. Showing Fingers: Say: Show me 3 fingers. Show me 8 fingers. 4. Telling Which Number is More: Say: Which is move, 9 or 62 Which is move, 42 or 28? S. Add and Subtract: Ask: (a) If you have 3 pen- nies ana spena 1 of them, how many have you left? (b) How many ave 3 apples and 4 apples? (c) Jack haa 9 marbles. He lost 3 of them. How many were left? Time Limits: 10 minutes for page of written computa- tions. 1 minute for counting 15 dots. 1 minute for reading all five numbers. 1 minute for showing fingers (both prob- lems). 1 minute for telling which is more (both problems). 1 minute for each of the three oral problems. Recording Oral Part: Counting dots—underline the last number cor- rectly counted and pointed to. Reading numbers, Show- ing fingers, Which is more, and Solving problems— underline numbers on form if correct; cross them out if incorrect. DO NOT SCORE WRITTEN PART. Cumul. Answer Key - Arithmetic, Level I Points Oral Part: Counts 15 dots, 1 point for each of the following: 1, 2-8, 4-5, 6-7, 8-9, 10-11, 12-13, 14-15.........eeee 8 Reads 5 numbers. . 5 pts. Fingers. . 2 pts.......... 15 Whichis more. . 2 pts. Answers to problems: 2 7 6... 20 000 55 APPENDIX II LEVEL II OF THE WIDE RANGE ACHIEVEMENT TEST FORM USED IN THE JUNIOR AND SENIOR HIGH SCHOOL STUDIES Personal Data The four lines at the top of the title page should be carefully filled out before the test is begun. This section provides necessary statistical information. The following uniform procedures should be observed in completing the blanks. Name: Print last name first, then first name and initials. Never assume that you know how to spell a name (not even Smith). Have S spell or write it for you on the line provided on the test form. A correct name may save much time in filing and finding records when needed. Birthdate: Example: 10-18-1955 for October 18,1955. M. F.: Encircle M for male; F for female. Chronological Age: List completed years and months up to age 15 yrs., 11 mos. For example, a child born on 7-21-1957 was 6 yrs., 3 mos. old on 11-15-1963. At 16 yrs. and above, list age in years only, using the year completed on the last birthday. A person born on 10-18-1943 was 19 yrs. old on 6-5-1963. School: Write down name of school attended at the time of the test. Grade: Enter the grade he is attending at the time of the examination in the case of school children. Date: Always record the date of the test. Example: 10-15-1960 for October 15, 1960. Examiner: Print name of person administering test. DO NOT COMPLETE ANY OTHER ITEMS. Now have the child write his name on the line be- low the little boxes on the first page. Reading Instructions Before administering this test, study the pronuncia- tion guide on page 5. The transcription symbols are those found in Webster's New Collegiate Dictionary. The E may use other standard dictionaries or the 56 symbols of the International Phonetic Association in learning to pronounce the words. Since this is primarily a reading testandnot a test of speech or diction, unusual pronunciations due to colloquialism, foreign accent, and defective articulation are accepted as correct. An incorrect answer is any misreading due to improper sequence of letter sounds, confusion of phonetic values, and misplaced accent. Always begin the administration with the word pronunciation test (74 words). Two copies of the test form may be used, one for S to read from, and one (with personal data filled out) for E to record on. Point to the first word "in'' and say: Look at each wora carefully ana say it aloua. Begin herve (point) and read the woras across the page so I can hear you. When you finish the first line, go on to the next line and then the next. In the case of young children (5 to 7 yrs.), each word should be pointed to with a pencil while S attempts to read. Time: 10 seconds per word. The reading part should be administered with as few interruptions as possible. Any clearcut response should be accepted and scored as either rightor wrong. The first time an error is made, S is asked to say the word again. His response is scored right, if he corrects himself on the second trial. From then on, the first response is scored as either right or wrong, unless S spontaneously corrects the error he has made. If the response is not clear, E may ask S to re- peat the word. The E should not intimate, by either motion or emotion, that he is dissatisfied with the answers. Spontaneous corrections are credited, but teaching, coaching, or questioning should be avoided. The reading speed may be controlled by E. Saying "next at the end of the time limit of 10 sec. is one way of controlling the rate of performance. Refusals to read within the time limits should not always be accepted as evidence of failure. If S hesitates or says "I don't know this word," E should encourage S to try the word anyway or ''take a guess'' at it. Testing Limits: 12 consecutive failures. Recording: a. Underline the first letter if the word is cor- rectly pronounced. Example: cat, block. b. Cross out the first letter of the mispronounced word. Example: ¢at, Flock. c. If S first mispronounces the word, then cor- rects his error, cross out the first letter and underline the second letter of the word. Example: fat, Hock. Score right. d. If S first pronounces the word correctly, then mispronounces it, underline the first letter and cross out the second letter of the word. Example: cit, bJock. Score wrong. On the reading test, some Ss tend to skim over the words or produce a response that sounds superficially correct. The E should be alert to these near successes and score them wrong, or ask S to repeat if no clear- cut decision can be made. Examples of such near successes are: = = v/v blsk for bl8k, Skskap for eskap, humiditi for vou s / : = = hiimiditi, kSntempori for kontémporari, akov for alkov, - — —/ uf kontémptius for kontémptiils, beni for bénin’, protuberans mil WX a v for protuberans, sentri’ figal for sentrifugal, abism4l wieV vv .. “ , for abizmal, susingkt for suksingkt; etc. Pre-Word Level: If S obtains a score of 10 points or less in the regular reading part, he should be asked to name the 13 capital letters printed above the word list and to name at least 2 letters in his name which he has written or printed on the line provided onthe first page of the test form. One point is assigned for each of the 2 letters in his name and the 13 letters to be identified. Level ll—Pronunciation Guide for Reading Test | in 2. WK inimmsimin milk 3. TRE, .ucummnserurywss tre 4. CY i. nmumienimanes sit i 5. animal.............. nd i mil 6. himself cc... commons him Bit” 7. between............. be twen’ 8. chin....cesmsmvmnins chin 9. Spliteeeeeecunnn... split 10, ZrOnt coisas grint 11. fOrMu...iesversvaners form 12. SIretCh...ocsmmiinsas strech 13. aboard.............. a bord’, 14, theory............... the’ ri 15. escape .............. és kap’ or is kap’ 16. grieve............... grev 2, 19. 20. 21. 22. 23, 24, 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42, 43. 44, 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72, 73. 74. contemporary..... kon tém’pd rér’l . toughen. ............ ttf €n contagious ......... kn ta’jus ethics............... Eh’ Tks image ........c..oue. fm’ ii triumph............. tri‘imf conspiracy......... k8n spir‘a st —_— a SW Sl iminate.. omen € limi nat rancid............... rén’s¥d tranquillity ........ trén kwilf of or tring kwil’¥ of deny......ce.eevvenn. de ni’ humiliate........... hii mil’Y at BICONE, 0s inuyrpsnres 41’kov Seal... com insure skold municipal .......... mil nis’{ pal desolate... ........ des’ 3 lit MOSAIC. covsvrursese mo 5 zat bibliography....... bib! iH og ra fi unanimous......... u {i nin’d mis decisive.... ........ dé si’siv contemptuous ..... kon temp’, th tis predatory .......... préd’a to r benigh...umsmes be nin’ deteriorate......... dé teri rat protuberance...... prd to/ber ins stratagem.......... strit’/a jem regime.............. ra zhem’ predilection........ pre‘d! 18k/shiin prevalence......... previa 1€ns irascible........... i rds} pb 1 or i- peculiarity......... pe ¢ uit 8r/f of abysmal....... ..... a biz ‘m8 pugilist ..oicvminees pa iy ist soliloquize......... so Ms 5 kwiz enigmatic .......... & ‘nig mark or eng mitdk centrifugal......... sén tried gl emaciated.......... & ma’shi at &d oligarchy....cvie ay gir kf covetousness...... kilv/& tls ns ingratiating........ in ge ‘shi at ing coercion............ kd Or ‘shiin . vehemence......... ved méns sepulcher,.... vie. sép/ll dl ker, longevity...... ..... 16n jevt o evanescence....... ev'a nds ns beneficent.......... bé nd % sént Subtlety....eovesnens sur’ Lu succinct ............ suk Sings beatify... «.coviswsns be at’ 6 fi regicidal.. ......... ré 1 { sd ¢:! BChigM.cvurvuusvunes siz'm heinous ............. ha ‘nis desuetude.......... dés’wé tad egregious .......... & gre’ills or &greff Us, y J misogyny........... mi s8jX nl or mI 887% ni ni internecine ....,... fn’ ter ne ‘stn synecdoche ........ si nek’ds k& ebullience.......... & pf ns 57 Arithmetic—Level II Instructions: Say: This is an avithmetic test. Turn to page 3 where it says Avithmetic, Level II ana look at the problems printed below the heavy line (hold test form up and point). I'a like to know how many of the problems on this page you can figure out. Look at each problem carefully to see what you ave supposed to do - add, subtract, multiply, or divide - and then pul down your answer in the space on or undev the lines. Should you wish to figure on the paper, you may use the empty spaces ov the margins to write on. First do the top row, then the second vow, then the third row, etc. The problems get more difficult as you go down the page. Don't spend too much time on any one prob- lem. You can skip a problem if it is too difficult for you, but do as many as you can one by one. You will have 10 minutes. Now, go ahead and do as many as you can. Time: 10 minutes for page of computations. DO NOT SCORE RESULTS. 58 Oral Arithmetic—Level II Any person obtaining less than five points in the written part must be given the oral part as follows: 1. Counting 15 dots: Point to the dots printed at the top of page 2 of the test form and say: Point with your finger and count these dots one by one beginning herve (S's left) ana going this way (moving to the right). Count them aloud so 1 can hear you and -telé me how many dots there are. (On top of page check once for correct count from 1 to 6 and again for correct count from 6 to 15.) 2. Reading numbers 3, 5, 6, 17, 41: Point to the numbers (printed upside down on page 2 of the test form) and say: Read these numbers. What is this ? (Pointing to 3) And this? Etc. 3. Solving three problems: Ask: (a) If you have three pennies and spend one of them, how many have you left? (b) How many are three apples and four apples? (c) Jack had nine marbles. He lost three of them. How many were left? (Record answers to questions a, b, and c at the top of the test form page.) 000 APPENDIX lI PERSONNEL PARTICIPATING IN THE DATA COLLECTION AND ANALYSIS Project Director K. Warner Schaie, Ph.D. Project Secretary Margaret D. Auberle Research Assistants Ronald Bone, A.B. Don Simons Claude Southerly, M.A. Barbara Stone, A.B. Field Testing Consultants Ruth Camp, M.A. (Morgantown, W. Va.) John Ivanoff, Ph.D. (Milwaukee, Wis.) Seymour Levitan, Ph.D. (Los Angeles, Calif.) Naomi A. Patterson, Ph.D. (Fort Collins, Colo.) Examiners MONONGALIA COUNTY, W, VA, Clay-Battelle Group John Clovis (in charge) Anne Barr Robert Beach Linda Bosley Virginia Bunner Carillon Copeland Margie Fox Patricia Lowry Felix Lunghi Fred Miller Michael Resetar William Sanders Kent Staggers William Sterling Bonnie Tennant Gladys White Huey Wilson Central Group Vivian Price (in charge) Margaret Brand Eleanor Collins Maxine Glover Eleanor Henry Katherine Reed Cheat Lake Group Harold Pickens (in charge) Patricia Baker Ronald Colovincenzo Thomas W. Cobun Mabel B. Harrah John Kapsaroff Jane Lemley Mary McGregor Elaine Shale William Spangler First Ward Group William F. Cunningham (in charge) Ethel Bagshaw Marguerite Franklin Julia Frum Daun Johnson Virginia Long Carol Mayle Thelma Morgan Dorotha Morris Ruth Morton Irene Mossburg Virginia Pixler Pauline Trickett Mildred Williams Westover Group Mrs. Sylva Fetty (in charge) Mrs. L. Douglas Curnutte Mrs. William Fournier Elizabeth J. Hall Anne Heiskell 59 Mrs. Claude Kemper Mrs. C. N. Rosenecker W. E. Spangler University High Group Michael Caruso Wayne G. Baker Allen R. Bryant Gloria J. Cunningham Janet D. Callahan Sara E. Logan David Loughrie William P. Hawley Lillie W. Morgan Richard N. Ryan Joseph P. Talerico Darrel E. Wood FORT COLLINS, COLO. Dick DeCook John Dyck Karen Gabbert Michael Gaynor Joel Gold Dal Hedlund Al Hinkle Bill Jones Anita Leighton Robert Leighton Larry Motzkus 60 Joan Muser Marilyn Rhodes James L. Sheard Lowell Wilson LOS ANGELES COUNTY, CALIF, Kerby Alvy Maureen Behrens G. William Dombhoff Carol Grieshaber Richard Mach Frank E. Webb MILWAUKEE COUNTY, WIS. Sara Derman Charles Droege D. Enders James Hanlon Margaret A. Hendricks Joanne Ivanoff Helen Kilgore Anne Nesbit Dianne Perone Lalia Peterson Lois Schaper Dennis Schrank S. B. Schultz Maribeth Schultz —o000—— %* U.S. GOVERNMENT PRINTING OFFICE : 1967 0—264-201 Series 1. Series 2. Series 3. Series 4. Series 10. Series 11. Series 12. Series 13. Series 20. Seyies 21. Series 22. OUTLINE OF REPORT SERIES FOR VITAL AND HEALTH STATISTICS Public Health Service Publication No. 1000 Programs and collection procedures.—Reports which describe the general programs of the National Center for Health Statistics and its offices and divisions, data collection methods used, definitions, and other material necessary for understanding the data. Data evaluation and methods research.—Studies of new statistical methodology including: experi- mental tests of new survey methods, studies of vital statistics collection methods, new analytical techniques, objective evaluations of reliability of collected data, contributions to statistical theory. Analytical studies.—Reports presenting analytical or interpretive studies based on vital and health statistics, carrying the analysis further than the expository types of reports in the other series. Documents and committee reports.— Final reports of major committees concerned with vital and health statistics, and documents such as recommended model vital registration laws and revised birth and death certificates. Data from the Health Intevview Survey.— Statistics on illness, accidental injuries, disability, use of hospital, medical, dental, and other services, and other health-related topics, based on data collected in a continuing national household interview survey. Data from the Health Examination Survey.—Data from direct examination, testing, and measure- ment of national samples of the population provide the basis for two types of reports: (1) estimates of the medically defined prevalence of specific diseases in the United States and the distributions of the population with respect to physical, physiological, and psychological characteristics; and (2) analysis of relationships among the various measurements without reference to an explicit finite universe of persons. Data from the Institutional Population Surveys.— Statistics relating to the health characteristics of persons in institutions, and on medical, nursing, and personal care received, based on national samples of establishments providing these services and samples of the residents or patients. Data from the Hospital Discharge Survey.—Statistics relating to discharged patients in short-stay hospitals, based on a sample of patient records in a national sample of hospitals. Data on movrtality.—Various statistics on mortality other than as included in annual or monthly reports—special analyses by cause of death, age, and other demographic variables, also geographic and time series analyses. Data on natality, marriage, and divorce. — Various statistics on natality, marriage, and divorce other than as included in annual or monthly reports—special analyses by demographic variables, also geographic and time series analyses, studies of fertility. Data from the National Natality and Mortality Surveys. —Statistics on characteristics of births and deaths not available from the vital records, based on sample surveys stemming from these records, including such topics as mortality by socioeconomic class, medical experience in the last year of life, characteristics of pregnancy, etc. For a listof titles of reports published in these series, write to: Office of Information National Center for Health Statistics U.S. Public Health Service Washington, D.C. 20201 NATIONAL . : CENTER Series 2 \ For HEALTH Number 25 ; STATISTICS — } TRA LEE AOA IT [2 U. S. DEPARTMENT OF Rg TM =o 0 eT: Sale NWA oR ANNs] Public Health Service Public Health Service Publication No. 1000-Series 2-No. 25 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C., 20402 - Price 40 cents NATIONAL CENTER| Series 2 For HEALTH STATISTICS | Number 25 VITALand HEALTH STATISTICS DATA EVALUATION AND METHODS RESEARCH Orthodontic Treatment Priority Index A description of the development of a system for studying malocclusion in population groups. Through the use of multiple regression equations the findings from the field examination procedure consisting of 10 objective components are weighted and summed to a single numberon a 10-point scale of case severity. A computer program for data processing is included. Washington, D.C. December 1967 U.S. DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE Public Health Service John W. Gardner William H. Stewart Secretary Surgeon General NATIONAL CENTER FOR HEALTH STATISTICS THEODORE D. WOOLSEY, DIRECTOR PHILIP S. LAWRENCE, Sc.D., Associate Director OSWALD K. SAGEN, Ph.D., Assistant Director for Health Statistics Development WALT R. SIMMONS, M.A., Assistant Director for Research and Scientific Development ALICE M. WATERHOUSE, M.D., Medical Consultant JAMES E. KELLY, D.D.S., Dental Advisor LOUIS R. STOLCIS, M.A., Executive Officer DONALD GREEN, [Information Officer DIVISION OF HEALTH EXAMINATION STATISTICS ARTHUR J. McDOWELL, Director JAMES T. BAIRD, JR., Chief, Analysis and Reports Rranch HENRY W. MILLER, Chief, Operations and Quality Control Branch PETER V. HAMILL, M.D., Medical Advisor LAWRENCE E. VAN KIRK, D.D.S., Dental Advisor Public Health Service Publication No. 1000-Series 2-No. 25 Library of Congress Catalog Card Number 67-6007 2 FOREWORD This study was begun as part of the Burling- ton Orthodontic Research Project of the University of Toronto and remains an essential part of the Canadian research. Interest in a treatment pri- ority index—an index of the handicapping extent of malocclusion—began at Burlington when the need arose to decide objectively whether pre- ventive treatment had reduced malocclusion be- low a level that might be considered of public health significance. A description of the proposed Treatment Priority Index (TPI) appeared in an annual report of the Burlington Orthodontic Re- search Centre, ! At about the same time, the Health Examina- tion Survey, a major program of the National Center for Health Statistics, was making plans for its second cycle of examinations. In Cycle I, a national probability sample of adults aged 18 through 79 years was examined with primary emphasis on cardiovascular disease, arthritis and rheumatism, and other chronic diseases. Cycle II would survey a sample of children 6-11 years old and would focus on factors related to growth and development. The dental examination would place special emphasis on the assessment of occlusion be- cause of its importance in this age group. But unlike most other areas of the dental examina- tion, no single survey assessment procedure had gained widespread acceptance or use. The items under consideration, when used together for the HES examination, would be a potpourri of time- tested clinical signs and symptoms, each in it- self capable of producing interesting and useful data, but missing by a considerable distance the most important and most needed statistic of all— an estimate of the extent and severity of maloc- clusion in the population. Now, the component parts of the TPI were quite similar to items already proposed for the Health Examination Survey and offered no new or untested measurement procedures. What was new was the potential ability of the index to summar- ize these heretofore disconnected clinical signs and symptoms into a single number on a 10-point scale of case severity and therefore make esti- mates of the severity of malocclusion in popula- tion groups. The value of such an index could not be overlooked and, to speed up the developmental work and the writing of a computer program for processing the results of Cycle II, financial assistance was provided by NCHS. Mr. Tavia Gordon, Assistant Chief, Division of Health Examination Statistics, participated by conveying the needs of the survey and, alongwith Dr. James E. Kelly, Dental Advisor to NCHS, aud Dr. Law- rence Van Kirk, Jr., Dental Advisor, Division of Health Examination Statistics, assisted Dr. Grainger through discussion and consultation. James E. Kelly, D.D.S. Dental Advisor National Center for Health Statistics Lawrence E. Van Kirk, Jr., D.D.S. Dental Advisor Division of Health Examination Statistics iii iv SYMBOLS Data VOL BB EYE ww vers mrss ws union woimiom:sewsiom am: sms: seme Category not applicable=-===-=ce-momcmanmn Quantity zZero=----=-=-===-meccmmmmemenna= Quantity more than O but less than 0.05---- Figure does not meet standards of reliability or precision----------cccee--= 0.0 CONTENTS Foreword ------=mmmmmm meme. Introduction -=-=-=-= mmm. Source of Data and Developmental Method--------------- Consideration of the Nature of Handicapping Malocclusion--------------- Establishment of a Scale for Expressing Case Severity Selection and Interrelations of Recording Items------- Development of Regression Equation for Estimating Treatment Priority SCOTS === mmm mmm mm me mmo Relation of Treatment Priority Index to Age------------- Computerized Mark Sense Examination System---------- Manual Field Examination Forms-----=-----cccooooo_- SUMMATY === === == == = =m mm mmm meee References ----- o-oo mmm. Detailed TableS----=-=-=-ccommmm mee. Appendix I, Definitions of Recording Items----------- Appendix II, Definition of Malocclusion Syndromes by Factor Analysis---- Appendix III, Development of Regression Equation------ Appendix 1V, IBM 7010 Fortran, Malocclusion Processor Page iq Ul DN = 14 15 15 18 18 20 29 39 42 vi This is a veport on the research development of a means of objectively assessing the degree of handicap due to malocclusion in terms of a Treatment Priovity Index. The work is based on the study of the inter- relationships of 10 manifestations of malocclusion as they occurved in 375 12-year-old childven with no history of ovthodontic treatment. The group constitutes a representative sample of children, primarily of Anglo-Saxon ovigin, from three Ontavio communities. The method was to define the natural groupings of manifestations which tended to occur jointly and which might be referred to as syndvomes. A judgment of the severity of the malocclusion for each child was obtained through divect examination by ovthodontic specialists. The highest val- ues in a 10-point scale indicated seveve malocclusion. Using multiple regression methods, formulas weve developed for estimating the judg- ment scoves from the objective measurements. The correlation between the calculated scove and the actual clinical judgment was comparable to that between two sets of clinical judgments. It is suggested that the index may be useful in epidemiological studies, as well as in initial screening of populations to determine the need for treatment while pro- viding a rough description of the case type. A fully computerized data-processing system and a manual form on which lo vecovdand calculate the Treatment Priovity Index ave provided. ORTHODONTIC TREATMENT PRIORITY INDEX R. M. Grainger, D.D.S., M.Sc.D., D.D.P.H., Faculty of Dentistry, Universily of Toronto INTRODUCTION The last decade has seen increasing interest in the development of indexes of occlusal status and many useful and interesting methods have been put forward.>7 No one method appears to be equally suitable for the use of epidemiologists, public health program planners, and clinicians. 8,9 Consequently it was felt that a renewed effort was needed to develop an assessment procedure that would objectively express the severity of maloc- clusion in clinically descriptive terms and, atthe same time, would be simple enough to be used by individuals without specialty training. The present interest in a treatment priority index, that is, an index of the degree of handi- capping malocclusion, began in connection with the Burlington Interceptive Orthodontic Research Project, 10 where the need arose to decide ob- jectively whether treatment had reduced the de- fect to below the level of public health significance. It was soon recognized thereafter that the same method would be useful for population surveys of the epidemiologic type and also as a screening device in public health programs. This study describes the development and use of a simple method of assessing the severity of the most common types of malocclusion and hence provides a means for ranking individuals according to their severity of malocclusion, their degree of handicap, or their priority for treatment. Although each ranking implies a different purpose, each quite obviously assesses the same thing. SOURCE OF DATA AND DEVELOPMENTAL METHOD A storehouse of invaluable records, particu- larly for a developmental study, is available at the Burlington Orthodontic Research Centre, Sets of dental casts are on file from a cross-sectional sample of children of the town at ages 3, 6, 8, 10, and 12 years. In each age group, 85 to 90 percent of the children at that age are included. In addition, for two groups serial dental casts were made annually for a period of 10 years. The 3-year-olds in the cross-sectional sample became the serial experimental group on which preventive ortho- dontic procedures would be performed as needed. The 6-year-olds became the serial control group. These unique records are valuable for study be- cause they are representative of all the types of occlusion in a typical community and also be- cause, for these children, there had been very little treatment that might obscure the natural patterns of malocclusion. Consideration of the Nature of Handicapping Malocclusion Strictly speaking, malocclusion is any dis- harmonious variation from the accepted or theo- retical normal arrangement of the teeth, But, in nature some degree of variation among individuals of a species is always present; hence the state- ment must be qualified as to the critical amount of variation which constitutes malocclusion. For the purpose of this study the ideal occlusion was taken to be the norm and the point from which variation is measured. It is not so much the amount of variation of linear measurements from their respective norms that causes malocclusion, but more importantly the inconsistent variation of parts. Thus it does not matter if all measurements of a faceare large compared with a set of skeletal norms; this merely means that the whole face is large. But when one measurement tends to be small while the restare larger than average, there is disharmony, and if the disharmonious part is closely related to the masticatory structure, there is a great likelihood of malocclusion. Neither does lack of complete harmonious conformity to norms necessarily constitute a malocclusion, as is shown in making dentures, where the production of minor variations to give a lifelike appearance is a fine art. How- ever, the degree of tolerated disharmony needs to be carefully determined for a specific popula- tion group if a realistic public health measure of handicapping defects is to be obtained. After careful consideration of what consti- tutes a handicapping anomaly from both profes- sional and lay standards, the Burlington project staff agreed upon the following as prerequisites for determining a handicap: 1. Unacceptable esthetics. Significant reduction in the masticatory func- tion. 3. A traumatic condition which predisposes to tissue destruction in the form of periodontal disease or caries. 4. Speech impairment. 5. Lack of stability so that the present occlusion will not be maintainable over a reasonable pe- riod of time. 6. In addition there exists a class of rare but gross, traumatic defects such as cleft palate, harelip, and pathological or surgical injuries which are unquestionably of very high treat- ment priority. Table A. Number and percent distribution of 375 12-year-old children, by clini- cally judged case severity scores: Bur- Malocclusion Severity Scale Interpretation 0 meee Virtually classic normal occlusion 1 Bl wre cme Minor manifestations and treatment need is slight 3 4 a Definite malocclusion but treatment elective 6 7 gl ----m-e--- Severe handicap, treatment highly desirable 9 10 ----------- Very severe handicap with treatment mandatory Figure |. Arbitrary scale for expressing case severity by means of a simple integer value be- tween zero and ten. lington, Brantford, and Orangeville, Ontario Percent Judge Number | distri- bution Total=-----===--== 375 100.0 Oemmmmmmm mmm mmm emo 22 59 lommmmm mmm eee me - 46 12.3 2mm mm mmm 68 18.1 Berm 72 19.2 tata 44 132.7 Smee mmm meee 40 10.7 Gmmmmm mmm meme em 36 9.6 Jom mmm mmm meme meee 29 747 Bom mmm meee 16 4.3 mma 2 0.5 10=---mmmmmrm mmm = - - For practical purposes, it was agreed that six conditions should be detectable either through a measurement or because of the obvious severity of the condition. Neither the costnor the difficulty of treatment would be considered in rating the handicap. Establishment of a Scale for Expressing Case Severity Figure 1 illustrates an arbitrary scale be- tween zero and ten that was selected as a means of expressing the degree of handicap or the priority of treatment which should be given, It was assumed that case severity is a continuum and thatno cut- off point existed below which treatment might be said with certainty to be unnecessary. The scale would express the degree and relative importance of the six conditions mentioned above as they oc- curred in a given individual. Theoretically, if enough trained personnel were available, it would be possible to undertake surveys of populations by simply recording judgments of the individual person's position on the scale. The high cost of employing orthodon- tists, even if they were available, makes this approach impractical. The alternative of mathe- matically estimating the judgments from objective observations became the principal goal of the study. In order to develop the estimating equations, a set of clinical judgments was needed. The dental casts of 203 12-year-old children in the Burling- ton collection and an additional 172 childreninthe nearby communities of Brantford and Orangeville were examined by members of the Burlington Proj- ect staff and the judgments recorded (table A). The clinical judgments of case severity are in no sense absolute. Rather, they are subject to considerable inter- and intra-examiner error. Figure 2 shows the degree of agreement between two orthodontists on the Burlington staff for 95 cases. The product moment correlation r is +.84; the average difference is 1.35; and the stand- ard deviation of the differences 1.40. Thus 19 out of 20 times, the examiners differed by as much as 2.8 points on a 10-point scale. However, a mathematically calculated estimate of case se- verity that differs no more from the judgment score of Orthodontist A than Orthodontist A dif- fers from Orthodontist B has certain advantages: (1) the priority estimate mathematically computed 10 9 z o - 8 ° x w > a7 —¢ 5 _— $e > w 2 }- OQ S 5 —do— > w Z a 4 So z a 3 ° oO => 3 8 2 wn So—o* 0 5° 0 | 2 3 5 6 7 8 °) 10 FIRST CLINICIAN'S JUDGMENT OF SEVERITY Figure 2. Relation between treatment priority judgments by two independent orthodontists for 95 cases from the Burlington cross-section control group of I2-year-old children. Table B. Distribution of 10 manifestations of malocclusion and normalized scores for 375 12-year-old children: Burlington, Brantford, and Orangeville, Ontario Upper Lower Congen- pp : : i ital 1 anterior | anterior Cverbite | Openbite esl Measurement over jet overjet Boss f T r T f 7 f T f T Ormmmemm cmc mm ————— 1.9] 27] 98.9 50 1:1 241 98.9] 50} 97.9 50 lommmemer remem mre mee 6.7] 34] 0.5 741 3.5 31 1.0 76 | 0.8 71 FE 26.4 | 421 0.5 78 46.9 | 44 - .. 1.3 75 3mm mmm ————— 24.3 | 49 =| ...|] 20.5| 60 - - TP EE EEE EE 16.3] 55 - 2.41 69 - - Somme me 0.8] 57 - 1.6 74 - - bmn mm mmm mmm mmm 7:51 62 - “| wee - - Jmmmmmm mmm mmm = 4.5] 66 - - - - 8mm mmm meme mmm meme 1.6] 70 - - - - mmm mmm meme meme me 2.4) 72 - - - - 10-=cmmm mmm emma 0.5] 76 - - - - Mean----====--mc=meemme mmm 3.54 0.02 2.01 0.01 0.04 Variance-----=-==-mme=mm=maa-- 3.83 0.03 0.77 0.01 0.06 Standard deviation------------ 1.96 0.16 0.87 0.10 0.25 : Posterior Posterior cross- Disto- Mesio- Gross. bite, aah _ clusion clusion te, maxillary ESplLace weagurement] maxillary to ment to bucca lingual f T f T f T f T f T mio sm 8 oe 61.3| 4597.9 5095.2 49 | 94.1| 49 50.4 | 43 se ELE i A a SA 10.4 | 54 0.3 71 1.6 68 2,9] 67] 11.2 52 mm am a a a a 16.8 59 1.0 23 1.3] 69 1.8 70 | 16.8 55 mc mm mm a cn om 3.2] 63] 0.8 771 0.5 71 0.8 751 4.3] 59 Liem mc con a a A 8.3] 67 =] ass 0.5 73] 0.3 79 6.4] 61 Dit mt 0 =| eee - 0.5 75 od J 2.9| 63 ly mom me i a ww - 0.3 79 - 2.9, 65 Tn ot - - -1 we - 1.6 | 67 Bnmmgietie 1 mui a = wt ———— w = = - 2d 70 ia - * = 0.3 75 10m mmm i nm nn a - = = - 0.5] 85 ME alms mmm wim m———————— 0.87 0.06 0.12 0.10 1.56 Variance === =m=m=mm == mm=mmmm=- 1.64 0.17 0.42 0.21 4.81 Standard deviation------------ 1.28 0.41 0.65 0.46 2.19 See Appendix I for definitions. f= frequency; T=normalized score calculated from areas in normal distribution X = 50, $='10 from the basic recordings would be reproducible, hence more useful for epidemiologic work, (2) less trained personnel could be used for examining, (3) the bias of the calculated score would become defined after a period of use and hence could be offset by knowledgeable interpretation, but the bias in subjective clinical assessments is not constant, thus not correctable, and (4) a calculated severity score could be useful as basic data in estimating the need and costs of treatment by a panel of ex- perts. Selection and Interrelations of Recording ltems The items to be observed either from casts ob Anterior openbite. Congenital absence of incisors. 6. Distoclusion as determined from buccal seg- > or direct clinical examination were selected be- cause of their bearing on the six points used to determine the degree of handicapping. The items were rigidly restricted to those that describe an occlusion anomaly, excluding factors bearing on etiology such as habits, or underlying measure- ments which are related to malocclusion but are not malocclusion per se, such as arch width. Moreover, a few manifestations of malocclusion such as midline diastema and slight asymmetry were rejected as being of little public health significance. Ten manifestations were chosen to be of primary importance. An eleventh re- cording item was included for special cases in which measurements seem inappropriate. The rare but severe defects such as cleft palate and other gross dento-facial anomalies would be re- ported here and automatically assigned the high- ments. 7. Mesioclusion as determined from buccal seg- ments. 8. Posterior crossbite with maxillary segment to buccal of normal cusp relation. 9. Posterior crossbite with maxillary segment to lingual of normal cusp relation. 10. 11. gross facial anomalies. Displacement of individual teeth. Cleft palate, traumatic conditions, and other The frequency distributions of the 10 record- ing items for the same 12-year-old children that were given judgment ratings for severity by the Burlington staff are shown in table B and in the detailed tables for other age groups with data from the dental examination of the National Center for Health Statistics' Health Exainination Survey, !l est case severity score, For definitions of these recording items, see Appendix I. 1. Upper anterior segment overjet. 2. Lower anterior segment over jet. 3. Overbite of upper anteriors over lower an- (It is important to notice that the distribution of measurements and scores was similar for the Burlington research data and theHESdata.) The average changes in the observations with age are, no doubt, anatomically real and indicate the need for slightly different standards or interpretations teriors. for different ages. Table C. Analysis of 105 pairs of replicated examinations (21 cases by 5 examiners) . S.D. of distri-|{ 95 percent Ce ene 5% bution of confidence possible range of Order of Recording Leen differences |range for SCOT og recording & between pairs |recording e ° relia- repro- bility Vzd?/2N ducibilicy Range Percent Anterior overjet----------- .969 +1.94 12 mm, 16.2 3 Anterior underjet---------- a-—n— err mm —— ——— Anterior overbite---------- 412 + .82| 5 points 16.4 4 Anterior openbite---------- .526 *1.06 5 mm. 21.2 6 Congenital missing incisor- i --- — iid — Distoclusion----==-=---==-= .572 +1.14| 4 points 28.5 8 Mesioclusion--==--memcmc-=-- .265 + .54| 4 points 13.5 1 Posterior crossbite, maxillary to buccal------- 1.28 +2.56| 10 teeth 25.6 7 Posterior crossbite, maxillary to lingual------ . 201 +1.40| 10 teeth 14.0 2 Tooth displacement--------- 1.86 +2.72| 15 points 18.1 5 S.D.—standard deviation. Table D. Classification of 126 cases having manifestations of malocclusion of a clini- cally significant degree: Burlington, Brantford, and Orangeville, Ontario Recording item and clinically significant degree Disto- Disto- clusions clusions . . with with Tio Mesioclusions overjet overjet Maxillary overjet, 5 mm. and over----====--- Vv Vv Vv V Mandibular overjet, 1 mm. and over--------- Vv Overbite, score 3 Or Over=-----eeeccccoea-- \ v Vv V Vv V Vv Vv . Openbite, 1 mm. and over------ccecmmccceaaa- Congenital incisor, score 1 or over--=-=----- Distoclusion, score 3 or over=------cecaea- Mesioclusion, score 3 or over=====---eceaaa- Posterior crossbite, maxillary to buccal, score 1 Or over=-----ccececcmmcccccceeeeem Posterior crossbite, maxillary to lingual, score 1 or over-----cememeomcemeeee eo Tooth displacement, score 4 or over-------- Vv Vv Observed frequency of combination-=--------- 141 40) 3) 1] 4 6] 15 1| 1 3 A Totals=====memc mecca 58 26 8 Some idea of the reproducibility of the re- cording items was obtained by analysis of repli- cated examination of 21 cases by five different orthodontists. This gave 105 pairs of replications for each of the 10 recording items. The results given in table C are, of course, directly applicable to the particular cases and examiners involved, but they do indicate where the most difficulty is encountered and where the most concentrated calibration effort would be needed when instructing new examiners. The most difficulty seemed to arise in recording distoclusion, posterior cross- bite, and anterior openbite, As a first exploration of the interrelation- ship of the first 10 recording items, a classifica- tion of the various combinations of defects of severity sufficient to be of some significance was tabulated. The critical severity levels for the various defects were determined completely arbi- trarily on an individual basis, and it is not sug- gested that these truly represent the levels of severity considered to be of public health signifi- Table D. Classification of 126 cases having manifestations of malocclusion of a clini- cally significant degree: Burlington, Brantford, and Orangeville, Ontario——Con. Neutroclusion Guevpine ’ Cross- displacements Very bites Congen- Condensed table ital incisor | ny | p2 | ML | M2 | Dis |0+0 | © 58 | 26 4 4 | 21 |11 2 21 11 1 cance. The levels used, as defined in table D, were: 1. Maxillary overjet------- 5 mm. and over 2. Mandibular overjet ----- 1 mm. andover 3. Overbite-e-mmccmccaaaa- score of 3 or over 4. Openbite-=cmecceecccana 1 mm. or over 5. Congenitally missing incisors---e--ecee-ca-- 1 or more 6. Distoclusion =-=-e=ceceu- score of 3or 4 7. Mesioclusion---==a-m--- score of 3or 4 8,9. Posterior crossbite----- count of 1 or more 10. Tooth displacement ----- score of 4 or over At the extreme right of table D is a condensation of the main blocks which correspond well with earlier work.!? It is unlikely that the severity judgments set by the clinicians would be directly related to any specific single measurement change because the clinicians were told to judge the cases without, as it were, performing a formal diagnosis. Nor can it be assumed that the grossest defect was the cause of the high judgment score because more moderate variation of another factor could con- ceivably be more important clinically. The judg- ments then must in each case apply to a specific combination of observations. In Appendix II, confirmation of the pattern of combinations (syndromes) originally proposed and presented in table D was undertaken by mul- tiple group factor analysis. Table of the appendix shows the results of analyzing a correlation ma- trix of phi coefficients. The general make up of the rotated factors was found similar to the syndromes in table D. The conclusion is that if the cases were separated into three sets—(l) the distoclusion group, (2) the mesioclusion group, and (3) the neutroclusion group—it would be reasonable to assume that the judgment expressed the severity of these syndromes and hence that multiple re- gression methods could be employed to estimate the score from the appropriate combinations of individual observations. estimating the clinician's Table E. Results of multiple regression calculations for judgment using squares of predictors for distoclusion, neutroclusion, and mesio- clusion cases T value for test Buccal segment re- Regres- of lation, recording Maan S.D sion regres- Srandopd Multiple yorbreel item,! and judgment 7 |coeffi- sion P a R ares score cient coeffi- r cient Distoclusion cases Upper overjet---=--===-= 21,12 | 22,30 052 9.76 .588 Upper overbite-------- 5.37 4.72 134 5.63 .386 Distal molar score----| 4,69 5.55 .067 2.96 215 P i : «795 1.58 osterior crossbite, maxillary to buccal--| 0.61 3.54 .115 3.88 «277 Displacement---=-=-=-=- 7.57 16.32 .023 3.40 «245 Judgment score-------- 3.96 2.31 “re ‘es PPP Neutroclusion cases Upper overjet=-----==-= 8.18 9.44 .088 13.71 .619 Overbite--===cccccaaax 3.42 3:03 .066 3.25 .184 Openbite--==c-eccmoann 0.01 0.10 «222 0.34 .020 Posterior crossbite, maxillary to buccal--| 0,15 1.70 .184 5.00 .276 .855 0:33 Posterior crossbite, maxillary to lingual-| 0,17 1.23 2251 4,79 «266 Displacement=---====== 3.62] 11,12 112 19.06 .739 Judgment score----==--- 1.76 1.99 ee oe A. Mesioclusion cases Lower overjet-=-----=-== 0.07 0.49 475 2.14 +256 Openbite------=ccce-n- 0.03 0.17 .100 0.14 +017 Mesial molar score---- 1.41 3.38 .260 6.20 .610 ] : .835 1.33 Posterior crossbite, maxillary to lingual-| 0.10 0.51 .403 1.63 .198 Displacement=---=-=-=-= 3.34 10.62 .059 5.75 .581 Judgment score-------- 1.97 1.56 cee iva -— All squared except judgment scores. S.D.—standard deviation. Development of Regression Equation for Estimating Treatment Priority Scores In the work of developing a computing equa- tion, squares of the individual recording items were used to provide better separation of thevery severe cases and to decrease the treatment pri- ority for cases with several minor defects that should not, even in combination, constitute a se- vere handicap. It was recognized that the weights or im- portance of items differ according to the combina- tion of other items present, To illustrate, 5 milli- meters of anterior overjet is not an extreme hdndicap in a neutroclusion case but in combina- tion with a slight distoclusion, it confirms the diagnosis of the syndrome and raises the impor- tance of the findings. Alternately, a mild upper anterior overjetoccurring ina case tending toward mesioclusion in the buccal segments may be an indication that the mesioclusion is of rather low severity and less likely to become worse. The regression of the measurements is only crudely linear with respect to clinical handicap— in fact, obvious break points exist. As examples, horizontal overjet becomes more critical at the point where the lower lip can reside behind the upper anterior teeth. Vertical overbite becomes suddenly severely handicapping when the lower teeth begin to impinge on the upper soft tissues. To compute the regression equations and de- termine the correlation of the judgment scores with the measurements, three subsets of records for the 12-year-old children were prepared ac- cording to the anteroposterior buccal segment relationship. In the distoclusion and mesioclusion sets, the groups used were all those individuals with some degree of distal or mesial molar de- fects, plus a few dozen very low priority scored cases in order to increase the effective range of variation. Under these circumstances the judg- ment score could be assumed to relate to the severity of the syndromes present. This would not have been so if the entire group of 375 had been used in each case, Calculations were made for the three molar relation groups using the BIMD # 6 program on the IBM 7094, Table E shows the resulting constants. While the results were generally good in terms of the multiple correlation coefficients which compare favorably with between-examiner correlation (fig. 2), deficiencies were apparent upon examination. First, the regression lines did not pass through the origin because the positive vertical intercept constants combined with posi- tive regression coefficients dictate that no esti- mate can fall below these values, Second, when the three equations were used in parallel on all cases, it was seen that the distoclusion equation gave estimates that were too low for the individ- uals with only one or two degrees of distoclusion. The neutroclusion formula provided a fairly suit- able estimate of these low-degree distoclusion cases but was much too high if used for full distoclusion dentitions. Third, there were a fair number of very gross discrepancies. Upon exami- nation it was evident these must be recording or judgment errors but they were included throughout the work because the source of the errors could not be verified. Fourth, the mesioclusion cases were rather poorly estimated as only a handful of seri- ous cases were available. The problem of the vertical axis intercept constant divergence from zero can be explained in terms of the factthat zero score had been taken as normal for the overjet and overbite. In the next round of calculations anterior overjet normal was to be taken as 2 millimeters and overbite as one-third so that all estimates might be reduced by this amount and some presumably to actual Zeros. The misfit of the distoclusion equation for cases with only slight buccal segment change and the value of the neutroclusion equation for these cases raised the problem of determining the buccal segment score cutoff point where one equation would be substituted by the other. What was needed was a single equation in which the weights for overjet, overbite, and displacement would grad- ually decrease as a function of the higher degrees of distoclusion. It was speculated also that the same need might exist while progressing from neu- troclusion to mesioclusion. Consequently, another set of multiple regression computations was done using seven groups instead of three in which all cases included in a group had the same type of buccal segment relation and the same degree of variation, For these calculations the data were rear- ranged so that anterior overjet-underjet and 1 Table F. Calculated regression coefficients and smooth values' for seven? buccal seg- ment relations Buccal segment relation Variable Distal | Distal | Distal | Distal | Neutral | Mesial | Mesial 4 3 2 1 1 Sample size--------cco---ooa 32 Ll 61 38 217 12 13 Y intercept----=---=-c--c--=- 4,59 3.87 3.57 1.97 0.12 2.29 0.63 Multiple R--=---=ccccmcnaonn .64 .68 .58 .69 .86 +359 .87 Judgment average------------= 5.78 5.09 4,98 3.68 1.99 2.33 3.69 Regression coefficients Anterior horizontal--------- +03 +05 .08 .13 .24 -.22 .08 Anterior vertical----------- .09 .28 .16 .04 +26 .20 +35 Posterior crossbite, maxillary to buccal-------- 14 «11 .10 .16 -2.61 14 Posterior crossbite, maxillary to lingual------- .96 -.76 .09 TT .27 -.34 + viy Displacement-=--=-c=cce=co=-u- .01 .01 +02 .03 11 .03 .62 Smoothed weights! Anterior horizontal®-------- 03 .05 .08 14 «24 14 .08 Anterior verticalf---------- .09 +11 213 .19 +23 .19 .15 Posterior crossbite, maxillary to buccal?------- 14 14 14 14 14 14 .14 Posterior crossbite, maxillary to lingual’ ------ .26 +26 .26 .26 .26 .26 .26 Displacement f-----cocccac--- .01 .02 .03 .06 .10 .06 .03 Y intercept?==-----umucmcun- 5.17 3.95 2.72 1.50 0.27 1.50 2.72 preliminary expressions from Appendix III used to construct smooth weights. 2No data were available for the obviously missing two higher sion groups. degrees of mesioclu- 3Horizontal (Y,- Y, — 2.0)%e “(14+ 0.53(Yg + v,)) vertical (v,- Y, - 1.0)2 e - (1.4 + 0.25(vg + Y¥,)) Crossbites were weighted averages. Displacement (v,,)? ec" (2.28 + 0.61Yg +.23Y,) TY intercept = .27 + 1.2(Y, + Y,) overbite-openbite would be continuous scales going from positive to negative. However pos- terior crossbite was not considered a continuum, because the two types can occur inthe same indi- vidual. Both types were observed in the same neutroclusion individuals but there was a higher tendency for buccal crossbite in the distoclusions and lingual crossbite in the mesioclusions. Conse- quently, these were left as separate items along 10 with the tooth displacement score. The buccal seg- ment score was dropped from the calculations be- cause all cases in the same group would have the same score and the contribution of this factor would be contained in the vertical intercept con- stant, Table F gives the results of the calculations for the seven groups and preliminary smoothed out regression coefficients. The smoothing out was Table G. Final smoothed weights according to buccal segment relations Buccal segment relation Variable Distal Rew Mesial 4 3 2 1 0 i 2 3 4 Horizontal componentl------u--- .07 .10 14 .19 wD +19 «14 .10 .07 Vertical component?------===--- 26) 31{ «39] .sL] .63] .31| .39] -31 24 Posterior crossbite, maxillary to buccald-------c--mmmmmmmao- 14 14 14 14 14 14 14 14 14 Posterior crossbite, maxillary to linguald-----ccmmeo-cmmmao- .26 .26 .26 .26 .26 .26 vob +26 .26 Displacement#----mcococcnacnnn- .01| .02| .o4| .07| .12| .08| .06| .05 .04 Constanti----==c-==co-ma-aao-== 5.07 | 3.95} 2.72| 1.50} 0.27 | 1.50 | 2.72 | 3.95 | 5.07 Horizontal (Y;-Y, 2yertical (v, = Y, _ 1.5)? e —-(.43 + -26(Yg + Y5)) 3Crossbites were weighted averages. ‘Displacement (¥,0)° e -(2.28 + .61Y¢ + .23Y,) _ 3.0)? a (1.34 + .32 (Yg + Y; 0) 5Vertical intercept constant 1.2(Y, + Y,) + .27 accomplished by expressing the regression co- efficients as exponential functions of the antero- posterior buccal segment relation. After further adjustments (see Appendix III), the equation below was derived and it was thought that a reasonable estimate of the clinical judgment was provided. J=027 +12(Y,+Y,) + ly, ~ Y _30)2 e” (1.34 + 32(Yg + Y,)) (v, A 1.5%" (43 +.26(Yg+Y,)) | 14 Yi 26 Y2+ Ye” (2.28 + .61(Yg + .23 Y,)) Table G gives the resulting regression co- efficients according to anteroposterior buccal segment relation. When the formula was used on 386 cases (375 12-year-old controls plus 11 additional mesioclusion cases which became available clini- cally), the correlation with the clinical judgment was + .795. The scatter diagram is shown in figure 3. Very few cases judged 7 or higher are not calcu- lated to be 6 or higher and variation in the lower end of the scale is judged to be of less consequence in that the important thing is for the calculated score to be reasonably linear and to be selective for the severe cases. There are a fair number of renegades, For most. cases either a gross clerical error or an obviously gross clinical judgment error must be the primary explanation. A second source of discrepancy was that clinical judgments of deep overbite cases (scores 4 and 5) were generally lower than those calculated and lower than could be accepted. Discussion of specific cases with the clinicians led to their agreement that deep overbite cases were likely underestimated when being considered from the esthetic viewpoint and that higher scores were more compatible. A third reason for discrepancies is that the esthetic handicap, for exampleofa cer- tain degree of crowding or of overjet, differs ac- cording to the facial type or lip fullness which may hide or emphasize the defect. It is not claimed at this stage that the weighting of the factors is 13 r= «795 12 L b 1 iQ we. — — Ss tn wm oe fe ia wn » w oc » > 3 9 : n * > q p= \ op op x 8 o + oc b r a. . b LJ - 7 » . 2 de a ope [| wl ¢ Ss o . 5 6 w * 4 x b r - I + qe p & ol 3 5 ] ] b < » J - S oo . 3 4 + 1 ] J a 3 b L < ae “le » . I oO - Ll q od LJ \ p oo 0 | 2 3 4q 5 6 7 8 9 10 CLINICAL JUDGMENTS Figure 3. Relation between calculated scores and clinically judged scores for 386 |2-year-old children. (375 controls + Il extra mesioclusion cases which became available clinically) TREATMENT PRIORITY INDEX under 4.5 NORMAL BEGIN OCCLUSION yes CONGENITA NCISORS no WHAT .1S DOMINANT DEFECT Displacement Horizontal Vertical Maxillary Maxillary INCISOR to to RELATION Underjet Overjet buccal lingual Overbite Openbite o> T— Wu -= 0 rl O00 oz x > a wm MACROGNATHIC SYNDROME IV RETROGNATHIC = o . A | Zw os = 3 no Oo o > x o a9 So @ az z Qo > > rr 2 ww wn = SYNDROME SYNDROME > > ax xh . od 0 = SO > w = oO x oO 2 >= wn J a n Ex o w 0? £0 oz OQ —- Figure 4 . Method used tosort cases into syndromes corresponding to factor analysis in appendix table |. perfect, but minor adjustments in the equation can be made in the light of experience. The final equation produces an objective measure of the handicap in the public health sense but does not indicate the syndrome or main maloc- clusion characteristics which cause the high score. In terms of a flow diagram, figure 4 gives the method of sorting cases according to the dominant defect and in a way that is compatible with the syndromes defined in appendix table I. Table H. Distribution of Treatment Priority Index for 375 12-year-old children, by syndrome: Burlington, Brantford, and Orangeville, Ontario Syndrome Ty Public | Tooth displacement Anterior Treatwent Prioricy Tots] | tools Pp normal ? Over- | Open- | Under- | Over- Bueal Linganl bite | bite jet jet I1 IV Vv III O-=-mmmmmm remem = 2.5 3.4 » we is .. .e lemme eee ma 38.6 51.8 oe 7 ag - ve 2mm mmm mmm meee 21.2 23.3 “s - "ow pa ‘ie 3mm meme 8.4 11.3 A. oe ce A. cos vu frome meme mm 8.0 5.1 wd 14.3 34.5 - - Se mmm mmm mmm meena m 5.5 _- - 14.3 17.2 - 50.0 31.4 Bmmmmm mmm meme LJ 33.3 7.1 24,1 25.0 50.0 23.5 Jmmm mmm meee mm 3.4 16,7 21.4 10.3 75.0 - 9.8 Brn mmm ee - 2.9 - 21.4 10.3 - - 9.8 mmm mmm meee 2.1 33.3 14.3 3.4 - - 7.8 10-=ccmmmm mre cme me 2.3 - 7.1 0.0 - - 11.8 Number---==-====-=-az- 375 269 6 14 29 4 2 51 Mean-=-=-==--c=-c--- 2.87 1.60 6.87 7.00 5.63 6.51 5.94 6.71 Standard deviation-- 2.34 0.82 .79 1.80 1.40 0.52 0.71 2.00 Standard error------ 0.12 0.05 #13 0.48 0.26 0.26 0.49 0.28 Table H shows the syndrome breakdown and priority rating for the 375 12-year-olds in the basic tabulation for survey findings. Included are the average Treatment Priority Index, its standard error, and a percent distribution of children by index score for the group as a whole and for sub- groups by syndrome. RELATION OF TREATMENT PRIORITY INDEX TO AGE The frequency distributions of the treatment priority scores and the scores sorted by syn- dromes are given in table J for the Burlington serial controls at ages 6, 9, 12, 14, and 16 years, and for some parents. The detailed tables give the individual syndrome summaries over the same ages, except syndrome V, prognathism, for which there were no cases. Thereis a general increasing average Treatment Priority Index of about one point from the youngest to the oldest age groups. The trend shown is, of course, influenced by the fact that the norms for some of the re- cording items differ considerably with age from 14 the values for the 12-year-olds used in con- structing the index (table B). The age trends seem, by inspection, to be most associated with syndromes I and VI which involve tooth displace- ment. For older ages the only comment needed is that the syndrome picture becomes less clear due to increasing tooth loss. For younger ages, three considerations are necessary: (1) until the permanent teeth are fully erupted, final tooth displacement syndromes are not observable, hence this syndrome picture must be incomplete; (2) there is a need to consider the severity of the malocclusion as it is currently present at the specific age; and (3) even more important is to attempt to project from the re- cordings how severe the anomalies will be at a later age if left untreated. This is the argument for expressing syndromes at early ages as esti- mates of the ultimate conditions at, say, age 12 years, It is better to use the same formula at all ages and to learn to interpret the results. It can be stated that because failure to detect syndromes at younger ages is due to failure of the syndromes to manifest themselves in terms of the 10 items Table J. Distribution of Treatment Priority Index by specified ages: Burlington serial control group Treatment Priority Index 6 years | 9 years | 12 years | 14 years | 16 years | Parents O0--=-=-mmmmmmmm mmm mmm em 4.6 5.0 4.7 65.9 0.0 5.6 lommmmmmm me mmmmm meme ee 32.0 21.5 23.2 28.1 25.6 25.0 2mm meme mmm ee 32.7 24.0 24.5 20.0 27.9 25.0 3mm mmm eee 13.4 20.7 18.0 17.5 14.0 22.2 fmmmmmm em mmm meee meme 5.6 15.3 11.6 13.1 11.6 2.8 Demmi m meme mmm m = 5.6 7.4 8.2 6.3 11.6 83 b-=mmmmmmm meme mem momo 2.3 3.3 4.3 37 2:3 0.0 Jemmmmmmeme mmm mmm meme 1.1 0.8 1.7 3.7 2.3 0.0 Bonner emememe-- 0.4 1.7 0.4 0.0 0.0 2.8 mmm mmm meee mmm 0.4 0.0 2.1 0.6 2.3 0.0 10------mmcmmmmm mm meme 1.8 0.4 1.3 0.0 2.3 8.3 Sample---=-=--=cmcmmomoaman- 284 242 238 160 43 36 Mean-=--==-===-==ccem=eomeaaa- 2.30 2.69 2.85 2.56 3.07 3.54 Standard deviation---------- 2.02 1.87 2.14 1.77 2.82 4.68 Standard error-------------- 0.12 0.12 0.14 0.14 0.43 0.78 recorded, the error isone of falsenegatives—i.e., the tendency will be to underrate malocclusion at early ages rather than to overrate it. The final answer to the problem of accurately predicting the degree and type of malocclusion at later ages from early signs must come through detailed study of each syndrome and probably will necessitate inclusion of additional etiologic obser - vations not included in this study of late clinical manifestations. As an example, a recent study by Scott ** has indicated the importance of the dis- crepancy in incisor width and the space from cus- pid to cuspid as an index of crowding. COMPUTERIZED MARK SENSE EXAMINATION SYSTEM A very convenient method of carrying out the Treatment Priority Index in field surveys has been developed, using an IBM mark sense card and a computer program, The card is shown in figure 5 and the computer program, written in IBM 7010 Fortran, given in Appendix IV, is used to compile the data. A summary of the computer output is given as table K. The IBM cards are marked with a special graphite pencil and the cards punched automatically on an IBM 514 punch in columns 1 to 20 after which the electronic computer com- pletely finishes the survey report. Instructions for setting up the cards for insertion into the computer are also given in Appendix IV. MANUAL FIELD EXAMINATION FORMS Figure 6 gives a form on which the examina- tions may be entered and the Treatment Priority Index calculated. It is used as follows: 1. Observe the first molar relation and place a check mark in the column heading which applies. 2. On the left hand margincircle the appro- priate measurement in millimeters for the horizontal incisor relation. Note that if this measurement is 2-4 millimeters it is considered normal with weight zero. 3, Also on the left hand margin circle the appropriate score for vertical incisor relation and for tooth displacement. An upper incisor overbite from zero to two- thirds is considered normal with weight zero. Also displacement scores zero and one are discarded with weight zero. 4, Find the appropriate weights for the first three items at the junction of the row and 15 Table K. Distribution of Treatment Priority Index for 375 12-year-old children, by occlusion group: Burlington, Brantford, and Orangeville, Ontario Treatment Priority Index Occlusion Average group Sample TPI S.E. 0 1 2 3 4 5 6 7 8 9 10 Total--- 375 2.8771 0.,12: 2,5] 38.64 2).2 8.4 8.0 S50 5.2 3.4 2.9 2.1 2.3 No malocclusion syndrome Normal oc- clusion-===-== 269 1.60] 0.05| 3.4] 51.8] 28.5 (11.3 5.1 - - - - - - Malocclusion syndrome Buccal dis- placement---- 6 6.87 0.73 - - -( 16.7 -1 33.3] 16.7 - 133.3 - Lingual dis- placement=---- 14 7.00] 0.48 - - - -| 14.3] 14.3 7.11 21.4] 21.4] 14.3 7-1 Overbite-=-=---- 29 5.63 0.26 - - - -1 34.5] 17.2] 24.1] 10.3] 10.3 3.4 0.0 Openbite=-==-==- 4 6.51] 0.26 - - - - - -1 25.0) 75.0 - - - Prognathism--- 2 5.94 0.49 - - - - - 1 50.0] 50.0 - - - Retrognathism- 51 6.71] 0.28 - - - - 5.91 31.41] 23.5 9.8 9.8 7.8 11.8 Congenital---- - “ee » wn - - - - - - - - - - S.E.-—standard error. coco] UPPER | Lower fovensire ; POST POST IDENTIFICATION OVERJET ovenuer [ crown [PTEL COMER | DUIS. | MESO | gre Xx BITE Dish AcEe AcE clball] MM wu | THIRDS MAX TO B | MAX TO L TeETY YEARS c0>c03c02lc03c0ciici ciclo ciclo cio oc0 kc0>c0oci>c(dciocidciociociocioc(ocD) cloc]oclok]lcloclac]loelorlocloe]loc] loc] kizcikineipkincit isco eime sels CCIE 2H 22C 23202) C2502 2oCI ICAI c3oc3oe3nc3oeloe ior ioe incitkeinc ine inci) 3c ioc iociac ioe ioc Ic ine ine ine ined CAAA ACIDIC Choc Cl CIscisdscicisCIScIscischetoe och choehoe Sesh Hoch 5ochoecsochHo) Ic5oehacsochaciacsehochrchochochocho chochH| CHIC 6] Ic6oChOchHOChICHOCH| “po ChIChOChHIAICHICHICH OCH OCHOCH OCH ICED, Coe Io c/o Ioc lola loco rset IoeIoe rock setae se oer | 89 Ic8oc8o8ocgocgoc8o Ic8oc8ac8oc8yc8ocgcgocrcgrcgrcgocgs <9 929 9oc9o9oc9 c95c99c9ocgotgoeg c9oc9ocHocIocIoci) ORTHODONTIC TREATMENT INDEX, FACULTY OF DENTISTRY, UNIVERSITY OF TORONTO IBM 7961 Figure 5. Mark sense card for field surveys of Treatment Priority Index. 16 NAME sescecsssssescsssscecsvene sesesessssssscscesecsesses DATEcsssessessoccsocscssce EXAMINERseoosossoeosenccsne National Center for Health Statistics and TREATMENT PRIORITY INDEX University of Toronto COMMUNITY eevesennne AGEtesesssreasesvoresrescans SEXeeeeoeosssascsecrnesosecnes CUBE Bla seus sunens sasereniioney (6) Distoclusion N (7) Mesioclusion e © 0 be! o u " 23 3 5 CASE TYPE t IS INDICATED FIRST MOLAR RELATION 2 8¢ | a 8 > 3 og 2g s Weights BY THE Choose appropriate column a ik” ad © o 9 @ ou @ DOMINANT I IP 4 3 G2 18a 3 DEFECT 1 ~ oo 1 ot de |e 1 @ ao % 1%) «© «a wg «a HH By ~ Sl © 0 — ~ wd | N = mm g ot | 2.0 | 3.4 | 5.4 |9.3 | 20+ | 9.3 | 50a |34 | 2.0 9 1.4 | 2.5 | 4.0 | 6.9 10+ | 6.9 | 4.0 |2.5 | 1.4 3 (1) Upper Overjet g | 1.0 | 1.8 | 2.8 [48 | 8.0 | 4.8 | 2.8 |1.8 | 1.0 SR 7 6 | 1.1 | 1.8 |30 5.1 | 3.0 | 1.8 |1.1 .6 o 6 I” 6 | 1.0 [1.7 2.9 | 1.7 | 1.0 8 “a a 5 5 2 3 4 .8 1.3 +5 “ 5 2 8 NORMAL Score 0 a .2 3 4 8 1.3 .8 4 3 .2 ; 0 “4 6 | 2.0 [1.7 2.9 | 1.7 | 1.0 i) “a 1 6 | 1.1 [1.8 | 3.0 5. | 3.0 | 1.8 |1.a 6 E (2) tower oerger 2 1.0 | 1.8 | 2.8 |4.8 8.0 | 4.8 | 2.8 | 1.8 | 1.0 rind E J 3 1.4 | 2.5 | 4.0 | 6.9 10+ | 6.9 | 4.0 | 2.5 | 1.4 8 3+ 2.0 3.4 S.4 9.3 10+ 9.3 S.4 3.4 2.0 — Bite | 2.9 | 3.8 | 4.8 | 6.2 8.0 | 6.2 | 4.8 | 3.8 | 4.9 § (3) Overvite in crom gjs, | 1,5 | 2.0 | 2.4 [3.2 | 41 | 3.2 | 2.4 |2.0 | Ls ane Ho 2/3-3/3| 5 7 9 [11 | 15 [1a .9 7 .5 Ag 25 0 +B Ih 9 1a | 15 2a .9 7 .5 SYNDROME IV : <2 1.5 | 2.0 | 2.2 [3.2 a1 | 3.2 | 2.4 [2.0 | 1.5 OPENBITE 2-4| 2.9 | 3.8 | 4.8 [6.2 8.0 | 6.2 | 4.8 | 3.8 | 2.9 4 bite in mm. . . . . (4) Openbite in mm a+ | 4.9 | 6.3 | 7.9 |10+ | 10+ [10+ | 7.9 | 6.3 | 4.9 Count teeth rotated 2 od 1 2 .3 “ oo 2 x a IS DISTOCLUSION AND/OR about 45° or dis=- 3 2 3 4 o7 1d 0 4 3 2 POSTERIOR CROSSBITE 8 placed about 2m.) 4 3 +5 «9 1.2 1.9 1.2 «9 5 3 MAX, TO BUCCAL 8 5 .5 8 | 1.2 [1.9 3.0 | 1.9 | 1.2 .8 .5 PRESENT B® Count teeth rotated 6 7 | 11 18 |2.8 | 43 [2.8 | 1.8 [1.1 2 — wo >45° or displaced 7 1.0 1.5 2.4 3.9 5.9 3.9 2.4 1.5 1.0 Synd: I a. VI § more than zmm)x 2 [] 8 | 1.3 | 1.9 | 3.1 | 4.9 7.7 | 4.9 | 3.1 | 1.9 | 1.3 Byuide a, 9 | 1.7 | 2.5 |41 |e.2 9.7 | 6.2 | 4.1 | 2.5 | 1.7 Bie ih brates 1 (0 y . ; : 0! . . . Total (0,1 no [] 9+ | 2.0 | 3.0 | 49 | 7.7 10+ | 7.7 | 49 | 3.0 | 2.0 Treuels Coane 9 score) CONSTANT 5,17 | 3.95 | 2.72 | 1.50 [0.27 [1.50 |2.72 [3.95 [s.17 SYNDROME VII (5) CONGENITALLY MISSING No. | o | 1 2 | 2+ | INCISORS werarr | o | 7 s | go | ©), MAX, TO no. Jo | 1] 2 3 | 4 |5 | 6 | 7 |8 |mre Sueg eons weIoHT| 0 | .1 | .6 | 1.3 12.2 [3.55.0 16.9 [9.0 | 10 8:8 MAX, TO No. jo | 1|2 | 3]|4|5 |6 | mre LINGUAL 258% WEIGHT] 0 | +3 | 1.0 | 2.3 14.2 |6.5 [9.4 | 10 OTHER DEFECT — — — — _ — — — mm Arbitrary Weight SUM OF WEIGHTS IS TREATMENT PRIORITY INDEX Figure 6. Manual examination and calculating form for deriving the Treatment Priority Index. column and enter this in the column to the right. 5. Transpose the constant for the correct column to the right, 6. Circle the correct scores for congenitally missing incisors and posterior crossbite, 7. Transpose the appropriate weights to the right hand column, 8. Add information on any other rare defects, such as cleft palates, that are observed. 9. If a rare defect has been observed that seriously modifies the index, add an arbitrary weight to ensure that the index will indicate its severity. 10. Add the weighting column to derive the Treatment Priority Index. The syndrome type is indicated by the dominant weight and the syndrome may be circled to be used as acrudedescription of the case. This does not constitute a diagnosis but does give an idea of the nature of the defect involved. The Treatment Priority Index derived from the manual form will not be exactly the same as that derived by solution of the full equation by computer due to rounding-off errors but may be equated for all practical purposes. The values in the manual form, figure 6, are the observed values multiplied by the appropriate regression coefficients. The constants are, of course, the vertical axis intercepts. Figure 6 sug- gests that a simplification might be possible. Many of the weights are negligible, hence, values corresponding are not worth recording. Also there is a level or point in the scale for most of the recording where clinical significance has been reached and recording much higher levels may be of only slight value, Thus future consideration may be given to recording the manifestations as dichotomies or at the most trichotomies, elimi- nating the labor of recording many relatively nor- mal conditions. SUMMARY This is a report on the research development of a means of objectively assessing the degree of handicap due to malocclusion in terms of a Treat- ment Priority Index. The work is based on the study of the interrelationships of 10 manifestations of malocclusion as they occurred in 375 12-year- old children with no history of orthodontic treat- ment. The group constitutes a representative sam- ple of children, principally of Anglo-Saxon origin, from three Ontario communities, The method was to define the natural group- ings of manifestations which tended to occur jointly and which might be referred to as syn- dromes and then, by regression methods, to de- termine weighting factors appropriate to each syndrome. A fully computerized data processing system and a manual form on which to record and calculate the Treatment Priority Index are pro- vided. The index may be useful in epidemiologic studies, as well as for initial screening of popula- tions to determine need for treatment while pro- viding a rough description of the case type. REFERENCES Burlington Orthodontic Research Centre: Progress Re- port No. 5. Faculty of Dentistry, University of Toronto, 1960. (mimeographed) Bjork, A., Krebs, A., and Solow, B. : Method for epide- miological registration of malocclusion. Acta Odont. Scand. 22:27-41, Feb. 1964. 3Canadinn Dental Association: Canadian Dental Survey Manual. A System for Recording and Statistical Analysis at the Community, Provincial and National Level. July 1959. Draker, H. L.: Handicapping labio-lingual deviations, a proposed index for public health purposes. 4m. J. Orthod. 46:295-305, Apr. 1960. SElsasser, W. A.: Studies of dento-facial morphology. 1. A simple instrument for appraising variations. Angle Orthod. 21(3):163-171, July 1951. Spoulton, D. R., and Aaronson, S. S.: Relationship be- tween occlusion and periodontal status. Am. J. Orthod. 47: 690-699, Sept. 1961. LLY Kirk, L. E., Jr., and Pennell, E. H.: Assessment of malocclusion in population groups. Am. J.Pub. Health 49(9): 1157-1163, Sept. 1959. 8 Grainger, R. M.: Indexing Handicapping Malocclusions. Working paper for meeting of Expert Committee on Dental Health, World Health Organization, Geneva, 1961. ® Grainger, R. M.: The Statistical Basis for International Dental Epidemiology. World Health Organization Document No. PA/98.65. Opopovich, F., and Grainger, R. M.: One community’s orthodontic problem, in R. E. Moyers and P. Jay, eds., Ortho- dontics in Mid-Century. St. Louis. C. V. Mosby Co., 1959. National Center for Health Statistics: preliminary un- published data. Grainger, R. M.: Interrelations of malocclusion syn- dromes.in Advances in Oral Biology (in press). Bgeott, L.: Personal communication. 14 Harman, H. H.: Modern Factor Analysis. Chicago. Uni- versity of Chicago Press, 1960. cL — C. C., and Van Voorhis, W. R.": Statistical Pro- cedures and Their Mathematical Bases. New York. McGraw- Hill Book Co., 1940. 16 Thurstone, L. L.: Multiple Factor Analysis, A Develop- ment and Expansion of the Vectors of Mind. Chicago. Univer- sity of Chicago Press, 1947. 17 BIMD: Computer programs. Division of Biostatistics, Department of Preventive Medicine & Public Health, School of Medicine, University of California, Los Angeles, Dec. 1961. 19 Table 20 10. 11. 12. 13. 14. 15. 16. DETAILED TABLES Distributions of upper anterior overjet measurements, by specified ages: Burl- ington Project and Health Examination Survey trial data=--==-==-eccocccececceaan Distributions of lower anterior overjet measurements, by specified ages: Burl- ington Project and Health Examination Survey trial data=-==--=-=--acecccccacccma-= Distributions of overbite measurements, by specified ages: Burlington Project and Health Examination Survey trial dat@===--=e-ececmceccocccecccmccacmccaccnaa- Distributions of openbite measurements, by specified ages: Burlington Project and Health Examination Survey trial data===-===mececmcmcccmecccacocccmcccccnm——- Distributions of number of congenitally missing incisors, by specified ages: Burlington Project and Health Examination Survey trial data=--=--====--eceecccaa- Distributions of distoclusion scores, by specified ages: Burlington Project and Health Examination Survey trial data-=--==meeecmcccmoocceaeceaaccccacccccnccaa- Distributions of mesioclusion scores, by specified ages: Burlington Project and Health Examination Survey trial dat@--==-=-m--cccmmccmm cc ccccmccccmcm emma Distributions of posterior crossbite, number of maxillary teeth to buccal, by specified ages: Burlington Project and Health Examination Survey trial data----- Distributions of posterior crossbite, number of maxillary teeth to lingual, by specified ages: Burlington Project and Health Examination Survey trial data-=--- Distributions of tooth displacement scores, by specified ages: Burlington Proj- ect and Health Examination Survey trial data@=----=--=ce-ccececccacecceacccnnaac- Distributions of Treatment Priority Index less than 4.5 (public health normal), by specified ages: Burlington untreated serial control group=----=-=-==--=ceec-a-- Distributions of Treatment Priority Index for syndrome I, buccal crossbite and displacement, by specified ages: Burlington untreated serial control group=----- Distributions of Treatment Priority Index for syndrome VI, lingual crossbite and displacement, by specified ages: Burlington untreated serial control group===-=--= Distributions of Treatment Priority Index for syndrome II, anterior overbite, by specified ages: Burlington untreated serial control group==--=-=-==e-cemcme-coa= Distributions of Treatment Priority Index for syndrome IV, anterior openbite, by specified ages: Burlington untreated serial control groupe===-===---ceec-eeccccca== Distributions of Treatment Priority Index for syndiome III, retrognathism, by specified ages: Burlington untreated serial control group===---=-=-eeeeacaecao-= Page 2: 21 22 22 23 22 24 24 25 25 26 26 27 27 28 28 Table 1. Distributions of upper anterior overjet measurements, by specified ages: Burlington Project and Health Examination Survey trial data 6 7 8 9 years 10 11 12 years 14 16 years | years | years years | years years | years Parents Oper Serial | Cross an to con- sec- over.je trol tion Burl. | HES! | HES! | Burl. | HES HES HES | Burl. | Burl. | Burl. | Burl. | Burl. 0 mm----- 5.2 - - 2,110.9 1.0 2:5 1.7 2.8 1.5 4.8 3.0 1 mm----- 22.6 2.3 5.6 6.6 3.6 4.7 7:1 8.0 5:2 26,1 4.8 24.0 2 mm----- 34.8 12.9 13.5] 24.4 14.9 16.4 14.0 23.1 23.51 32.9 28.6 28.0 3 mm----- 27.1 16.1] 24.3 31.41 24.317 25.3) 27.7 26.5 23.0 16.1 30.9 32.0 4 mm----- 10.1 2.7 19.1 17.31] 20.2 | 22.8 23.8 21.0 19,7 11.27 23.8 5.0 5 mm----- 2.9 6.8 10.5 12.4 10.3 | 13.9 10.5 10.9 1:5 4.3 7.1 3.0 6 mm----- 1.4 3.5 8.2 3:37 6.1 7.9 5.6 5.5 8.0 1.2 4.8 4.0 7 mm----- 0.3 2.5 4.5 3:31 5.3 4.0 4.4 2.5 4.7 0.6 - - 8 mm----- 0.3 1.3 2.2 0.4 2.0 2.5 1.0 0.4 1.4 - - 1.0 9 mm----- 0.3 1.3 - 0.41 0.7 0.7 2.2 0.4 4.2 - - 1.0 10 mm----- - 1.3 0.7 - 0.3 0.2 0.7 - - - - 1.0 11 mm----- - - 0.1 0.41 0.5 0.2 0.2 - - - - - 12 mm----- = - - - 0.2 - - - - - - 13 mm----- - 2,5 - - 0.2 - 0.2 - - - - - Sample---- 302 394 267 247 | 603 402 408 238 213 160 43 36 Mean------ 2.441 2.27 3.40 3.34 3.48 3.80 3.69 3.29 3.68 2.17] 3.14 2.62 Variance-- 1.54| 8.94| 4.33| 2.46] 4.71| 3.10 3.84 2.46 4.08 1.93| 1.44 2.82 Standard devia- tion----- 1.24 2,997 2.08 1.5371 2.17 1.76 1.96 1.57 2.02 1.39 1.20 1.68 lypper anterior overjet was not examined for primary dentitions in HES data. Table 2. Distributions of lower anterior overjet measurements, by specified ages: Burlington Project and Health Examination Survey trial data 6 7 8 9 years 10 11 12 years 14 16 Parents ears | years| years ears | years r Lower ¥ > ¥ y y Serial | Cross yesrs: [years anterior con- sec- overjet trol tion Burl. | HES' | HES' |Burl.| HES| HES| HES| Burl.| Burl.| Burl.| Burl.| Burl. 0 mm.----| 99.9 | 99.2 | 99.3 | 98.6] 99.2| 99.3| 99.7 99,2 98.6 99.3 | 97.7 100.0 1 om.---- 0.3 0.8 0.7 0.8 0.7 0,2 0.2 0.8 0.9 0.6 2.3 - 2 mm.---- 0.3 - - 0.4 - 0.2 - - 0.5 - - - 3 mm.---- - - - - - 0.2 - - - - - - Sample---- 302 394 267 247 | 603 402 408 238 213 160 43 36 Mean------ 0.01 0.0 { 0.01 | 0.02] 0.01| 0.01 0.0 0.01 0,02| 0,01 | 0,02 - Variance-- (0.014 0.0 |0.006 [0.026 (0.006 | 0.029 0.0f 0.008| 0.029] 0.006 |0.023 or Standard devia- tion----- 0.120 0.0 (0.083 [0.155(0.083| 0.167 0.0| 0.089] 0.170] 0,077 {0.151 . Lower anterior overjet was not examined for primary dentitions in HES data. 2] Table 3. Distributions of overbite measurements, by specified ages: Examination Survey trial data Burlington Project and Health 6 7 8 9 years 10 11 12 years 14 16 years | years | years years | years years | years Parents . Serial Cross Overbite CON SEC trol tion Burl, | HES! | HES! | Burl, | HES HES HES | Burl, |Burl. | Burl, |Burl. | Burl, Om=mmccmca- 10,0 - - 2,0 14.4 4,7 6.4 2,5 - 0,8 2,0 6.1 lemmm——ean 62.4 25.3 29,5 10.1] 24.4 24.3 24,0 36.5 23.9 49.7 52,4 70,7 YR ttl 17.6 23.5 | 42,6 42.4 | 46,7 55.3 52.9 47.5 50.2 39,2 38.1 22.2 Jemma 92.3 5.8 12.3 14.5] 13.4 14.9 14.7 13.0 24,4 9.2 7.1 - femme em 0.3 0.5 0.4 - 0.7 0.5 1.7 0.4 - 1.8 2.4 1.0 Semeccceam 0.3 0.5 - - - - 0.2 - 1.4 - - - Sample-=-= 302 394 267 247 603 402 408 238 213 160 43 36 Mean-=-=-- 1,28 0,9 1,53 1,38] 1,61 1,82 1,82 1,72 2.03 1,63 1.60 1.19 Variance-- | 0,656 | 1,040 | 0, 846 1,21 10,846 | 0,563 | 0,689 0,548 0, 608 0, 533 0,518 0, 348 Standard devia- tion----- 0,810 1,02 | 0,915 1,100,920 | 0,754 | 0,834 0,739 | 0.780 | 0,727 | 0,718 0,585 loverbite was not recorded for primary dentitions in HES data. Table 4, Distributions of openbite measurements, by specified ages: Examination Survey trial data Burlington Project and Health 6 7 8 9 years 10 11 12 years 14 16 years | years | years years | years years | years Parents Serial Cross Openbite cone sec trol tion Burl, HES HES Burl, | HES HES HES Burl, Burl, Burl, | Burl, | Burl, 0 mm=====- 92,3 96.0 95,0 98,2 95.7 95,2 96,0 98.3 100.0 99.3 97.7 95.0 1 mMee———— 5.6 33 3.0 1.6 3.0 3.2 3.4 t.7 - - 2.3 2,0 2 MM==-=-- 2,0 0.8 1.1 - 0.2 1:2 0,2 - - - - 2,0 3 mMMe-ee-- = = 0.7 - 0.5 - 0.2 - - 0.€ = - Sample-==-- 302 394 267 247 603 402 408 238 213 160 43 36 Mean------ 0.10 0,05 0,07 0,02 0,05 0,06 0,04 0,02 - 0.01 0,02 0,06 Variance--| 0,130] O, 063 0,137] 0, 017 0, 084 0, 078 0,058 0,017 eve | 0,053 0, 023 0.096 Standard devia- tion--=-= 0.356| 0,251 | 0.365] 0,125/0,285| 0,277 | 0,243 0,129 ... | 0,233 0,151 0.312 22 Table 5. Distributions of number of congenitally missing incisors, by specified ages: Burlington Proj- : ect and Health Examination Survey trial data 6 years 7 8 9 years 10 11 12 years 14 16 years | years years | years years | years Parents Number of congenitally Serial | Cross missing con- sec- incisors! trol | tion Burl. | HES| HES HES Burl. | HES | HES HES | Burl. Burl. | Burl. | Burl. | Burl. O-=ccmmemmm- 99.6 | ow -—- LE 9.4{ ~~~ -—— -—— 99.6 | 97.7| 99.3] 100.0 97.9 l-memmmmmmm- 0.3] ~== -——— -——— 0.4) --- -—— -—— 0.4 0.9 0.6 - 2.3 2mm | —— -— = -| --- -—— -— - 1.4 - - - Yt —— ny — wim —_— "| === Eo i - % w ” - Sample------ 302 247 238 213 160 43 36 Mean-------- 0.0 0.0 0.0 0,38] 0.01 = 0.02 Variance---- 0.0 0.0 0.0{ 0. 063 | 0.006 0.023 Standard deviation-- 0.0 0.0 0.0] 0.250 | 0.077 wi 0.151 lBurlington counts derived using radiographs. Table 6. Distributions of distoclusion scores, by specified ages: Burlington Project and Health Examination Survey trial data 6 years 7 8 9 years 10 11 12 years 14 16 years | years years | years years | years Parents Disto- Serial| Cross 21utien con- sec- ¢ trol tion Burl. HES HES HES Burl. HES HES HES Burl. Burl. | Burl. | Burl. | Burl. 0---====-- 71.6 77.2 64.5 61.7 50.1] 64.5] 62.2| 60.0 61.3 56.8 68.2) 65.1 78.4 l--mmmem- 13.8 6.2 12.1 10.8 20.8 11.4 11.4 15.2 14.7 10.8 11.6 18.6 13.5 2--mmmmm- 10.1 12.6 17.5 20.9 22.2 16.0 20.1 14.7 17.2 21.1 13.4 11.6 8.1 CE 3.0 1.3 2.5 3.7 5.3 3.6 3:5 3:7 2.9 2.8 4.3 2.3 - a 1.3 2.7 3.0 2.6 1.6 4.0 - 6.4 3.8 8.4 2.4 2.3 - Sample--- 302 373 394 267 247 603 402 408 238 213 160 43 36 Mean----- 0.48 0.46 | 0.67 0.74] 0.88 0.70| 0.62 0.81 0.73] 0.95| 0.61] 0.58 0.30 Variance- | 0.792 | 0.922 | 1.082 | 1.166 | 1.061 | 1.232] 0.865 | 1.464 1.704| 1.664 | 1.188 | 0.903 0.372 Standard devia- tion----| 0.893 0.955 1.04 1.08 1.03 1,11) 0.926] 1.21 1.31 1.29 1.09] 0.947 0.610 21 Table 7. Distributions of mesioclusion scores, by specified ages: Burlington Project and Health Examination Survey trial data 6 years 7 8 9 years 10 11 12 years 14 16 years | years years | years years | years Parents Mesioclusion Serial | Cross score con- sec- trol tion Burl. HES HES HES Burl. HES HES HES Burl. Burl. | Burl. | Burl. | Burl. Ommmmmm meee 92.1 80.4 90.6 95.0 95.7 93.7 91.8 92.9 92.9 99.1 88.9 86.0 97.3 lommmmmmmeeeem 5.4 4.0 3.8 1:5 3.6 2.8 bo 3.4 4.6 - 4.9 9.3 2,7 2mmmmmmm meee 2.0 11.8 3.8 2.6 0.4 2,0 1.2 2.5 2.5 0.9 5.5 4.7 - 3mm mem 0.3 1.6 1.0 0.4 - 0.8 1.0 0.5 - - 0.6 - - femme - 2.1 0.5 0.4 - 0.2 1.0 0.7 - - - - - Sample-------- 302 373 394 267 247 603 402 408 238 213 160 43 36 Mean---=------- 0.10 0.41 0.16 0.10 0.04 0.10 0.15| 0.13 0.10 0.02 | 0.18 0.19 0.03 Variance------ 0.152 | 0.828 | 0.548 | 0.212 | 0.053 | 0.203 | 0.336 | 0.281 | 0.137 | 0.036 | 0.292] 0.250 0.026 Standard devia- tion--------- 0.391) 0.912 | 0.735] 0.459 | 0.226 | 0.452 | 0.584 | 0.526 0.370 | 0.190 | 0.542 | 0.497 0.162 Table 8. Distributions of posterior crossbite, number of maxillary teeth to buccal, by specified ages: Burlington Project and Health Examination Survey trial data | + Posterior 6 years 7 8 9 years 10 11 12 years 14 16 Parents crossbite, years | years years | years years | years number of maxillary Serial | Cross teeth to con- sec- buccal trol tion Burl. HES HES HES Burl. HES HES HES Burl. Burl. | Burl. | Burl. | Burl. 97.5 99.2 99.2 99.1 94.9 98.0 98.0 95.6 92.9 95.8 22.0 86.0 97.9 0.7 0.5 0.3 0.4 3.6 0.7 1.0 1.7 3.3 0.9 4.3 9.3 1.0 1.0 0.3 0.3 0.4 0.8 0.3 0.2 1.5 2.5 1.4 2.4 4.7 1.0 0.7 - - - 0.4 0.2 0.2 0.7 0.8 0.9 0.6 - - - - - - - 0.3 0.2 0.5 - - 0.6 - - - - - - - - - - - 0.53 - - - - - - - - - - - - 0..5 - - - Sample-------- 302 373 394 267 247 603 402 408 238 213 160 43 36 Mean----==--=== 0.05 0.01| 0.01 0.01| 0.06 0.03 | 0,03] 0.09 0.112 | 0,117 | 0.13 | 0.19 0.03 Variance------ 0.109 | 0.017 | 0.014 | 0.020 | 0.102 | 0.084 | 0.068 | 0.212 | 0.203 | 0.423 | 0.270 | 0.250 0.048 Standard devia- tion-==-e-=--- 0.330 | 0.130 | 0.122 | 0.141 | 0.318 | 0.291 | 0.260 | 0.463 | 0.447 | 0.650 | 0.524 | 0.497 0.222 24 Table 9. Distributions of posterior crossbite, number of maxillary teeth to lingual, by specified ages: Burlington Project and Health Examination Survey trial data 6 years 7 8 9 years 10 1 12 years 14 16 PEEL Posterior years | years years | years years | years ents crossbite, number of Serial | Cross maxillary con- sec- teeth to trol tion lingual Burl HES HES HES | Burl. HES HES HES | Burl. Burl. | Burl. | Burl. | Burl. 96.2 | 97.3| 91.6] 93.5| 98.3| 89.8| 93.5| 90.7 92.4 | 93.0 | 94.4 93.2 97.9 0.7 0.8 4.0 2.6 2.8 4.1 2.5 4.2 4.2 3.8 37 2:3 2.0 1.3 0.3 1.3 1.1 2.0 2.5 1,7 2.2 2:1 1:9 1.2 2.3 - 1.7 0.8 1.8 1.5 1.6 1.8 1.0 1.5 0.8 1.4 0.6 - - - 0.3 0.5 - - 0.8 0.2 0.7 - - - - - - - - 0.4 - 0.2 0.5 0.2 0.4 - - - - - 0.3 0.5 0.7 - 0.2 - - - - - - - - 0.3 - - - - cd 0.2 - - - 2.3 - - - - - - 0.1 0.2 0.2 - - - - - Sample-------- 302 373 394 267 247 603 402 408 238 213 160 43 36 Mean---------- 0.08 0.09| 0.17] 0.16| 0.12] 0.21] 0.14| 0.20 0.13 | 0.12] 0.08 | 0.23 0.02 Variance------ 0.212] 0.397] 0.490 | 0.533 | 0.240] 0.578 | 0.449 | 0.624 | 0.281 | 0.230 [ 0.137 | 1.188 0.020 Standard deviation----| 0.456 | 0.625] 0.700 | 0.733 | 0.489| 0.760 | 0.673 | 0.787 | 0.533 | 0.480 | 0.366 1.09 0.141 Table 10. Distributions of tooth displacement scores, by specified ages: Burlington Project and Health Examination Survey trial data 6 7 8 9 years 10 11 12 years 14 16 Parents years | years | years years | years years | years Tooth displacement Serial | Cross score con- sec- trol tion Burl. HES HES | Burl. HES HES HES | Burl. Burl. | Burl. | Burl. | Burl. 46.9| 63.0 60.0| 24.6| 63.6 | 64.7| 56.8 23.5) 37.6] 17.2) 18.2 2.0 19,1 12.1} 15.0 19.9 11.5] 10.0} 12.7 20.2| 16.4] 13.5] 13.6 4.0 20.4 13.7] 14.6] 33.3] 10.0 10.0 9.6 18.5| 17.4 24.5| 22.7 27.3 8.4 5.8 3.0 18.3 7.0 5.0 5.4 13.4 4.6 19.0 | 20.4 26.3 4.0 3.8 3.8 8.0 5.0 3.7 6.6 9.2 9,91 11.6 6.8 21.2 1.0 0.3 1+5 1.6 13 1.7 3.2 7:1 4.2 8.0 6.8 6.1 - 0.5 0.4 2.1 0.8 1.7 1.7 3.8 3.3 3.1 6.8 6.1 - 0.3 0.8 1.6 0.5 0.5 1.0 2.1 2.4 1.8 2.3 5.1 - - 0.4 0.5 - 1.0 1.0 0.8 3:3 - - 2.0 - - - 0.5 0.2 0.7 0.2 0.8 0.5 1.2 2.3 - - 0.3 - - - 0.5 1.0 0.4 0.5 - - - - - - - - - 0.2 - - - - - - - - - - 0.2 - - - - - - - - - - - - 0.2 - - - - - - - 0.4 - - - - - - - - - Sample--=-====c=c=c--- 302 394 267 247 603 402 408 238 213 160 43 36 Mean------=-====o-==-=- 1.06| 0.82| 0.95| 2.14) 0.89 1.04 | 1.33 2,25] 1.97 | 2.48 | 2.59 3.41 Variance 1.538 | 1.850 | 3.497 | 2.560 | 2.190 | 3.803 | 4.666 | 4.326 | 5.290 | 3.534 | 4.452 2.789 Standard deviation-----==-=--- 1.24] 1.36 1.87 | 1.60 1.48 1.95] 2.16 2,08 2.30 1.88 | 2.11 1.67 25 Table ll. Distributions of Treatment Priority Index less than 4.5 (public health normal), by specified ages: Burlington untreated serial control group Age nearest birthday Treatment Priority Index Parents 6 9 12 14 16 years | years | years | years | years 0 £0 0.5m mmmmm meme meme eee 5.3 6.1 5.9 8.6 0.0 6.9 0.5 £0 1.5mm mmm mmm meee eee - 37.1] 26.4] 29.2 | 35.2| 33.3 31.0 1.5 £0 2.5mm mmm mee ee eee 38.0 | 29.4 | 30.8 | 25.0| 36.4 31.0 2.5 £0 3.5m mmm meee 15.5] 25.4 22.7] 21.9 18.2 27.6 3.5 £0 4.5mm meme em eee 4.1] 12.7 11.4 9.4] 12.1 3.4 Number in sample 284 242 238 160 43 36 Number of cases===-=-ccm comme - 245 197 185 128 33 29 Percent of cases of sample--===--ccmcmmmmcaaaaoo_- 86.2 81.4 79.3 80.0 76.7 80.5 Average TPI---mcme momo occa 1.68 | 2.05( 1.98 1.83] 1.96 1,92 Standard error--------- oem eaoa- 0.06 | 0.08| 0.08 0.09] 0.18 0.19 Table 12. Distributions of Treatment Priority Index for syndrome I, buccal crossbite and dis- placement, by specified ages: Burlington untreated serial control group Age nearest birthday Treatment Priority Index Parents 6 9 12 14 16 years | years | years | years | years 4.5 £0 545mm mmm mmm mmm meee eee - - 14.3 25.0 = 5.5 £0 6oS== mmm mmm mmm emma ame 100.0 | 100.0 | 28.6 | 50.0 | 100.0 66.7 6.5 £O 7.5mm mmm mmm mm - - -| 25.0 - - 7.5 tO 8,5--mmm=mmmmemmmmcmee ecm eee, —————— » - - - ® - 8.5 £0 9. 5mmmm mmm meme me eee - - - - - 33.3 9.5 to 1045 moon mm mm meee eee - -| 42.9 = = - EE ale - = 14.3 = - # Number in sample-=-====-cocmmmm meme - 284 242 238 160 43 36 Number of cases==-=-c-cocommmm a -- 1 2 7 4 1 3 Percent of cases of sample--------ceccmmomaoooaoooano 0.4 0.8 3.0 2.5 2.3 8.3 Average TPI---=cmo como coe eee . 5.09 7.76 | 5.00 we ® 5.62 Standard error--=--=---cemcmm eee eee wis 0.36 1.08 | 0.44 ’% 0.96 Table 13, Distributions of Treatment Priority Index for syndrome VI, placement, by specified ages: Burlington untreated serial control group lingual crossbite and dis= Age nearest birthday Treatment Priority Index Parents 6 9 12 14 16 years| years| years| years| years 4.5 tO 5.,5-===m=memmmccmmcemce sms e ese eee ee ———————— 100,0( 33.3] 12,5 20,0 - - 5.5 tO 6,5======mmmmemmmmcmmcemmm ese ee — em ee —————————— -| 50,0| 50,0| 20.0 - - 6.5 tO 7.5-==m=memmmemmmmmmccemeemcmc emcee m eee ——— -| 16.7 -| 20,0 - - 7.5 tOo 8,5-=====-memmcmmmcceeeecmeeee emcee m—— eee ———— - - -| 40.0 - - 845 £0 9. 5=m=mmm mm mm mm mm mmm mae - -| 12.5 - - - 9.5 to 10.,5========mmmrecmcmcccc mem ccm em ——— - -| 12.5 - - - 10.54===-emmmemmmcmmcmmmcccccmmmem emcee cece m———— - -| 12.5 -| 100.0 - Number in sample-------ceccocccccccccccmce cme 284 242 238 160 43 36 Number of caSeS=mm==-emmmcmcccccccccccmmcccecmenen—— 1 6 8 5 1 - Percent of cases of sample--=c=ccccmccmccammencecnnan= 0.4 2,5 3.4 3.1 2.3 - Average TPl---emmcmcmmcmcccccmmcccccccccccccccceec—a= id 4,77 6,52 5.93 we vias Standard error-----e-eeemmesmccecmmmcccceeeececeeeme———— cin 0.25 0,78 0,56 > en Table 14, Distributions of Treatment Priority Index for syndrome II, anterior overbite, by speci= fied ages: Burlington untreated serial control group Age nearest birthday Treatment Priority Index Parents 6 9 12 14 16 years | years | years | years | years 56S £0! F050 mm mmm mmm sw smim sm sm 0 22,2 44.4) 28.6| 42.9 - - 5.5 £0 6457" mmm emma 44,4) 55.6 | 42.9] 28,6 - - 6.5 tO 745 mmmmmmmmemee meee 22.2 -| 14.3 - - - 745 £0 8.5=m=mmmmmmememmmmm——————————————————— - -| 14.3] 14.3 - - 8.5 £0 9,5==m==mmmmmmmmmmemcemmemeeem neem ———————— - - < - - - D5 CO 10 Summmmmmm mm mmm mime rm om sr a a 0 11.1 = -| 14.3] 100.0 - 10 Shem mmm mm mm ee - - - - - - Number in sample--e--e-cccceccccccccecccccmcccnnnnaan 284 242 238 160 43 36 Number of caseS===c-memecmccmcmcmaccccaccccmcnconoaonn= 9 9 7 7 1 - Percent of cases of sample=ce--e-cceccceccccccaccnean 3.2 3.7 3.0 4.4 2.3 Average TPleeeeeemcmemceccccecececcccccocemcmmemenann 5.37| 4.63 | 5.30] 5.49 ces cor Standard error--e=-e=e-eceeececmccsscscssccsc-mece-e--== 0.44| 0.18] 0.37] 0.69 rp wen 27 Table 15. Distributions of Treatment Priority Index for syndrome IV, anterior openbite, by speci- fied ages: Burlington untreated serial control group Age nearest birthday Treatment Priority Index Parents 6 9 12 14 16 years | years | years | years | years 505 £0 640mm mmm mmm eee eee 35.7 -1 33.3 - = 645 LO 74 mr mmm mmm ee eee 28.5 - - - - - 745 LO 845mm mmm mmm mee meee eee em 7.1 -| 66.7 - - - 8.5 LO 9, 5mmmemunmenae mmm m an - - - - - - 945 £0 10.5 === mm mmm mmm meme m - - - - - - LO Sm mm mmm mm ee eee m 28.6 - - - - 100.0 Number in sample----===-me cco mm ome 284 242 238 160 43 36 Number of cases===-==c-ocmom meee 14 - 3 - - J. Percent of cases of sample----=-cccemmcmoocmmaocaaan 4.0 - 1.3 - - 2.7 Average TPI--------m oom cee een 7.38 6.33 ao . RN Standard error-==-------o oom eeeeaem 0.87 0.46 . . Table 16. Distributions of Treatment Priority Index for syndrome III, retrognathism, by specified ages: Burlington untreated serial control group Age nearest birthday Treatment Priority Index Parents 6 9 12 14 16 years | years | years | years | years 4.5 EO 545=mm==mmm mmm mmm eee meen 23:1] 27.3 5.9 33.3| 20.0 - 505 £0 64 5mmmm mmm mmm mmm eee eee eee 38.5 | 22.7| 41.2] 11.1] 80.0 - 6.5 LO 74 mmmmm mmm mmm mm eee eee eee e 7.7 22.7) 41.2) 33.3 - - 7¢5 £0 84 mmmm mmm mm me me eee - 15.4 4.5 5.9 22.2 - - 845 £0 94 mmm mmm mm me meee 7.7| 18.2 - - - - 9¢5 t0 10 5mm mmm meme meee - - - - - = 10 St mmm mm me eee ee ee 7.7 4.5 5.9 - - 100.0 Number in sample=--===c=ccmcmmm meee ae 284 242 238 160 43 36 Number of casesS=-==-c-mmcmm meee 13 22 17 9 5 2 Percent of cases of sample-------cccmmmmmmmmaoaaaaoa- 4.6 9.1 7.3 5.6 11.6 5.6 Average TPI---== momo em eee eee oe 5.84 6.00 5.74| 5.46 | 4.78 12,07 Standard error=-=-=-- === meee 0.53] 0.48 0.37] 0.37| 0.18 2.17 28 APPENDIX I. For use in the Burlington Research Project, se- lected manifestations of malocclusion were used asre- cording items. For initial recording purposes, these items were defined as follows. Hovizontal incisor relationship.—The maxillary incisors may protrude beyond the lower incisors in the horizontal direction (Item 1. overjet) or vice versa (Item 2. underjet). Have the subject close together his posterior teeth; place a ruler horizontally at the mid- line against the labial central incisor surfaces of the less protrusive ‘arch and measure to the outside of the incisor tip (fig. I). If the central incisors are not in similar anterior position, take an average judgment. oO — Mandibular Over jet (under jet) He Over jet Method of measuring horizontal incisor rela- tionship. Figure I. Vertical incisor velationship.— Excluding cases where the incisors are not close to being completely erupted, have the subject close his posterior teeth and observe whether the central incisors overlap on the vertical direction (Item 3. overbite) or if they are still spaced (Item 4, openbite). Note the amount of overbite according to the horizontal position of the incisor tip of the most prominent arch, Judge the amount of open- bite if present in millimeters (fig. II). Congenitally missing permanent incisors (Item 5).— This could not be determined with certainty without radiographs but if at age 12 the teeth were obviously not in sight, the count was recorded. Anteroposterior buccal segment relation (Item 6. distoclusion, neutroclusion, or Item 7. mesioclusion).— Describe the anteroposterior position of the lower teeth to the upper teeth, paying particular attention tothe re- lation of the upper and lower first permanent molars and, if present, the deciduous second molars (fig. III). DEFINITIONS OF RECORDING ITEMS Score overbite 5 i ET IR Passed lower gingival margin. overbite 4 WHHD OE