DOCTORAL DISSERTATION SERIES Tm F 4 S M I OF THE &UATION BETWEI cmm MENTAL M b PERSONAUT1/ TlkITS ANb MT/NdS OF MUSICAL mams tiiTHOB JOHN CfffclSftPtfEA C 00i.il/ UNIVERSlTY ^ y/£4K S F jf£ COLL, DATE DEGREE £d'D. PUBLICATION NO. ^ T % » UNIVERSITY MICROFILMS " ANN ARBOR • M I C H I G A N A STUDY OF THE RELATION BETWEEN CERTAIN MENTAL AND PERSONALITY TRAITS AND RATINGS OF MUSICAL ABILITIES By John Christopher Cooley A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science In partial fulfillment of the requirements for the degree of DOCTOR OF EDUCATION School of Education 1952 John Christopher Cooley candidate for the degree of Doctor of Education Final examination, July 24, 1952, 10:00 A.M., Room 328b, Morrill Hall. Dissertation: A Study of the Relation between Certain Mental and Personality Traits and Ratings of Musical Abilities Outline of Studied Major subject: Minor subject: Education Music Biographical Items Born, January 24, 1924, Inman, South Carolina Undergraduate Studies, Furman University, 1941-45 Graduate studies, Michigan State College, 1945-47, cont. 1947-52 Experience: Graduate Assistant, Michigan State College, 1945-47, Music Director, Webbervllle High School, 1947-48, Music Director, James Couzens High School, Bath, Michigan, 1948-49, Music Director, East Lansing Public Schools, 1950-52. Member of Phi Mu Alpha, Hand and Torch Scholastic Fraternity ACKN GhLEDGEAiEJiTS i The author wishes to express his sincere appx*eclation to Dr. Milosh liuntyan, under whose stimulating guidance this investigation was undertaken ahd. completed. He is also Indebted to Dr. William R. Sur Tor his i invaluable suggestions and encouragement throughout the study. Grateful acknowledgement is also due Dr. C. V. Millard, Dr. Carl Gross, and Dr. Guy Hill for their helpful suggestions In organizing the study and in writing the manuscript. The writer deeply appreciates the help of Dr. Roy Underwood whose cooperation in the data collection and unfailing interest throughout the study made It possible to undertake and complete this investigation. Grateful acknowledgement is made to the meir.Lers of the Music Department Faculty for their suggestions and Invaluable cooperation In collecting data for the study. The author is also Indebted to Dr. William A. Mann and Dr. Owen Reed for their suggestions and cooperation In data collection. The Investigator extends his thanks to Hr. Leisenrlng of the Registrar's Office and to Mr. Robert Jackson of the Board of Examiners for permission to use data contained In the study. TABLE OF CONTENTS CHAPTER I. PAGE THE BACKGROUND OF THE PROBLEM AND THE TECHNIQUE OF SOLUTION ............................ 1 The Background or the Problem • • » • • • • • Talent testing 1 .......................... The German studies 2 ...................... 6 Summary Statement Indicating the Validity or the Technique or S o l u t i o n ........ The test or scholastic a b i l i t y ......... 8 The test or reading a b i l i t y ............. 9 The test or p e r s o n a l i t y ................ 10 The test or musical t a l e n t ............. 10 The musical ability ratings ............. Applied music grades II. 8 .................. 11 • Direction or the s t u d y .............. 13 SUMMARY OF PERTINENT LITERATURE.......... 14 The American S t u d i e s .......... Validation studies 14 ...................... Musical aptitude and mental traits • Musical perrormance and mental traits The German Studies 12 15 • • • 19 ... 21 ................ 22 Relation or muslcallty to other types oT a b i l i t y ............................ 22 PAGE Relation of musicality to ability to perform • 24 S u m m a r y ........................................ . 25 III. THE GROUP S T U D I E D ................................... 28 Description of the Group Studied .............. 28 Conditions of Data Collection.....................34 S u m m a r y ........................................... 36 IV. SOURCES OF D A T A ..................................... 38 The American Council on Education Psychological 39 Examination................................ Studies of validity and reliability............ 43 Summary .............. 44 The Cooperative Reading Comprehension Tests . . . 45 The Bernreuter Personality Inventory 48 Summary .......... ..................................... 54 The Seashore Measures of Musical Talent . . . . . 55 The 1939 revision of the Seashore Measures of musical talents ...................59 Musical ability ratings V. • 63 M u s i c a l i t y ................................ 65 Ability to sight-read 65 .................. Ability to perform . . . . . . 65 Applied Music Grades . . . . . . . . 66 SUMMARY OF STUDY D A T A ............................... 68 Abbreviations for Tests . . . . . . . 68 The American Council on Education Psychological Examination R e s u l t s .......... 69 PA OS The Cooperative Reading Comprehension Test Results • 79 The Bernreuter Personality Inventory Results • • • • • • • .................. 90 Results of the Musical Ability Ratings and Applied Music Grades • * ................ 124 Summary of R e s u l t s ........................ 136 The cooperative reading comprehension test 137 The Bernreuter personality Inventory r e s u l t s .......................... The Seashore results • .................. 139 The musical ability rating results • . • . 140 ........ 141 SIGNIFICANCE OF THE STUDY D A T A .............. 142 The applied music grade results VI. 138 Significance of the Age Results . . . . . . 142 The Significance of the Cooperative Reading Comprehension Tc^t R e s u l t s .............. 143 The Significance of the Bernreuter Personality Inventory Results . . . . . . The Significance of the Seashore Results • • 144 146 Relation of Music Ability Ratings to Test Scores • • • • • • • . . « • . • VII. ........ 147 The test profile of the g r o u p .......... 149 CONCLUSIONS AND INTERPRETATIONS.............. 153 Conclusions and Interpretations of the Study 154 PAGE Relation of intelligence and reading ability tom u s i c a l i t y ................... Relation of Seashore scores to musicality • 155 The relation of personality to musicality • 155 Summary VIII. ............................... EDUCATIONALIMPLICATIONSAND SUGGESTED RESEARCH Educational Implications .......... 156 157 • • • • • 157 Suggested Further Research • • • • • • • . • • 155 B I B L I O G R A P H Y ......................................... APPENDIX 154 159 LIST OF TABLES TABLE PAGE I* Average ACE Scores for the Four Classes . . • II. III. Average ACE Scores for the Curricular Groups 77 Average ACE Scores for the Male ana Female Groups V. 76 Average ACE Scores for the Instrumental G r o u p s ................................... IV. 75 ............................. 75 Correlation Coefficients Found between ACE Scores and the Musical Ability Ratings and Applied MusicGrades VI. VII. VIII. ................. Average CRCT Scores for the Four Classes • • 87 Average CRCT Scores for the Curricular Groups QQ Average CRCT Scores for the Instrumental G r o u p s ...................... IX. 89 Average CRCT Scores for the Male and Female G r o u p s ................................... X. 80 91 Correlation Coefficients Found between CRCT Scores and the Musical Ability Ratings and the AppliedMusicG r a d e s .............. XI. XII. XIII. 92 Average EPI Scores for the Four Classes . . • 102 Average BPI Scores for the Curricular Groups 103 Average BPI Scores for the Instrumental Groups............. ....................... 105 TABLE XXV* XV. PAGE Average BPI Scores for Male and Female Groups • 106 Correlation Coefficients Found between BPI Scores and the Musical Ability Ratings and the Applied Music G r a d e s ............... XVI* Average Seashore Scores for the Four Classes XVII* Average Seashore Scores for the Instrumental G r o u p s ..................................... XVIII. ............................... 118 119 1C1 Average Seashore Scores for the Male and Female Groups ............................... XX. • Average Seashore Scores for the Curricular Groups XIX. 107 122 Correlation Coefficients Found between Seashore Scores end the Musical Ability Ratings and the Applied Music Grades XXI. .... AverageRatings of Musical Ability for the Four C l a s s e s ............................... XXII. ..................... 132 Average Ratings of Z&islcal Ability for the Male and Female G r o u p s ..................... XXV. 130 Average Ratings of Musical Ability for the Instrumental Groups XXIV. 129 Average Ratings of Musical Ability for the Curricular Groups ........................... XXIII. 123 133 Applied Music Grade Point Avereges for the Four Classes ............................. 135 TABLE XXVI* PAGE Applied t*U3ic Grade Point Averages for the Curricril&r Groups . . . . . . . . . . . . . . XXVII. 135 Applied lAiSic Grade Point Averages for the Instrumental G r o u p s .......................... 136 LIST OF GRAPHS GRAPH 1 PAGE Distribution or Ages at Which Music Study was B e g u n ...................................... .. . 30 2 Distribution or Students by C u r r i c u l a ........... • 32 3 Distribution or Students by Major Instrument G r o u p i n g s .............. 33 4 Distribution or Theory Grades •• 5 Distribution or ACE-T S c o r e s .................... 70 6 Distribution or ACE-L Scores .................... 72 7 Distribution or ACE-Q S c o r e s .................. . . 73 8 Distribution or CRCT-T Scores.. .................. 81 9 Distribution or CRCT-V Scores ................. 83 10 Distribution or CRCT-R S c o r e s .................. 84 11 Distribution or CRCT-C S c o r e s .................. 85 12 Distribution or Bernreuter Personality Inventory .............. Bl-N S c o r e s ........ . ........................ 13 • • • • • 96 Distribution of Bernreuter Personality Inventory B4-D Scores 16 95 Distribution of Bernreuter Personality Inventory B3-I S c o r e s .................................. 15 93 Distribution or Bernreuter Personality Inventory B2-S Scores 14 35 .............................. 97 Distribution or Bernreuter Personality Inventory Fl-C S c o r e s .................................. 99 GRAPH 17 PAGE Distribution of Bernreuter Personality’ Inventory F2-S S c o r e s ................................ 100 18 Distribution of Seashore Pitch Scores ......... 109 19 Distribution of Seashore Loudness Scores • , . • 110 20 Distribution of Seashore Time S c o r e s ........... Ill 21 Distribution of Seashore Timbre Scores ......... 112 22 Distribution of Seashore Rhythm Scores . • . • . 114 23 Distribution of Seashore Tonal Memory Scores • * 115 24 Distribution ofSeashore Six-Test Average Scores 117 25 Distribution of 125 26 Distribution of Average Ratings of SightReading Ability 27 Average Ratings of l&islcallty • .......... 127 Distribution of Average Ratings of Performance A b i l i t y .................................... 28 Distribution of Applied Music Grade Averages • 128 134 CHAPTER I THE BACKGROUND OF THE PROBLEM AND T1IE TECHNIQUE OF SOLUTION This chapter consists of (1) a summary statement of the Background from which the problem arises, and (2) a summary statement of the method used in the study to solve the problem. The first section, concerned with the background of the problem, points up two ways by which musical talent has been studied. This section includes a brief contrast, drawn between ”atomistic” and "gestalt” approaches to the study of musical talent. The second section of this chapter, concerned with the technique of the study, presents a summary statement indicating the validity of the technique of solution. This section provides information concerning the sources of data utilized in the problem. I. THE BACKGROUND OF THE PROBLEM Within the past three decades, research in music education has outlined, broadly, two points of view as to the nature of musical talent. The existence of these two points of view has, in large part, been due to differences In approach and technique of study. 8 Talent testing* One approach to the study or musical talent has been that or talent testing, and has been concerned mainly viith group tests or musical talent. The work or Carl E. Seashore is widely known in this country ror Cl) its early contribution to musical talent testing and (8) its laboratory research into the nature or musical talent. Seashore selected certain psychological and aural Tactcrs oT musical talent, and constructed tests designed to measure these factors. His approach to musical talent test con­ struction has been, largely, that or an acoustical analysis or the components or musical sound. Geashore^ constructed tests based on the rollowing characteristics or musical sound: pitch, loudness, timbre, and time. Other or Seashore's measures were tests or rhythm, tonal memory, and consonance. CInformation as to the validity and reliability of the Seashore measures will be round in Chapter Four or this study. Other pertinent information will be round in Chapter Two.) The Seashore tests have employed a "llmenal" type or construction. Utilising the characteristics or musical sound listed above, these tests have been based on the subject's ability to make various kinds or discriminations between pairs 1 Carl E. Seashore, Measures of Musical Talent. NY: Columbia Phonograph Co., 1919 and Carl E. Seashore, Measures or Musical Talents. 1939 revision, Camden, N. Jersey1 ftCA Victor, 1939. 3 or sonar stimuli. The purpose of the tests has been to determine the least distinguishable difference that the subject could identify; that is, to determine the threshold of aural sensitivity to these sonar characteristics. Test items of the Seashore battery are of a "same-different" order, requiring a response to acoustical phenomena. For example, the pitch test presents items consisting of paired tones which have small differences in vibration frequency (pitch). The subject is asked for a response involving his ability to determine which tone is the higher (or lower) in pitch (vibration frequency ). Kwalwasser is also widely known in this country for his work in musical talent test construction. tests of Kwalwasser and Dykema p The musical capacity are similax* to those of Seashore, and employ a similar approach to the study of musical talent. The technique of study in both the Seashore and KwalwasserDykema tests has been an ’•atomistic" one. This technique is based on an analysis of musical talent into many independent factors. Seashore comments as follows regarding this point: No matter how many members we have in a battery, each member remains a specific measure; that 1.3, the technical validation must be made in terms of the thing measured in each one. The more members of basic sig­ nificance we have in the battery, the larger command 2 Jacob Kwalwasser and Peter Dykema, Kwalwasser-Dvkeroa Music Tests: Manual of Directions. NYs Carl Fischer, Inc., 1930. or the situation it should give. This is what we have called the specific theory of measurement as opposed to the omnihiiB theory which aims to validate the battery against the total situation in musical performance. It is clear, then, that these tests are based on an analysis of musical talent into discrete elements. These two tests (those of Seashore and Kwalwasser-Dykema) or musical talent have been widely used in this country (1) in studies of group differences and (2) in correlation studies of mental abilities. Regarding the results of these studies, Bienstock states: The status of testing and guidance In music is beginning to emerge as a subject worthy of intensive effort by both psychologists and musicians. The results, however*, ax*e far from conclusive at the present time.4 This point of view, representative of Seashore, Kwalwasser anu others,5 has tended to show that these aspects of musical talent (measured by these musical talent tests) are Independent of other mental abilities, and that the presence of unusual musical talent does not presuppose the coincidence of other specific abilities, or of unusual generul ability. 3 Joseph Saetveit, Carl Seashore, and Don Lewis, The Revision o£_ tip* Seashore Mfe&Pureff,, o£ jfelSlqal TftkSflS.. Iowa City: Universityof Iowa fress, 1939, p. 48. ^ Sylvia F. Bienstock, "A Review of Recent Studies on Musical Aptitude, * Journal of Educational Psychology,# 33: 440, 1942. 5 Ibio.. 33: 427-442. 5 £ Seashore T a I ftrvt points out, however, that The Measures of qaT do not furnish a single, ail-inclusive index to musical ability, but that each score is an item in the musical profile. (For further evidence and discussion of these points and those to follow, see Chapter Two, "Summary of Research" and Chapter Four, "Sources of Data.") American studies, concerned with the Seashore and Kwalwasser-Dykema tests, have indicated little correlation between what is termed "musical talent" and such factors as general intelligence and personality facets Ctraits ). studies These it»ve pointed up the relation of these musical talent tests to functional musicality. These studies have not shown that these musical talent tests can separate positively, the 7 musical person from the non-musical person. The musical talent tests of Seashore and of Kwalwasser-Dykema do tend to make this musical-non-musical discrimination negatively, however. It can be stated, then, that these aural abilities are of great importance, as criteria of musical talent, but apparently, do not encompass all important aspects of musical talent. 6 Carl S. Seashore, Joseph G. Caetveit, and Don Lewis, Manual Instructions Aflna &b£L fiftflffhftEtt. Measures of Musical Talent. New Jersey: RCA Victor Div., Radio Corp. of America, 1939, 19 pp. £*'®e Chapter Four. 6 The German studies. An entirely different approach to the study of musical talent has been used by certain German rtGestalt 1st s. **S In contrast to the atomistic technique, employed in the studies referred to previously, these German studies have employed a "gestalt" approach, thereby taking c . . much broader, more functional criteria of musicality."' Based on functional criteria of musical talent, these studies found: (1) that the number of people with special talents tends to decrease with intelligence, and (2) that there Is a high relationship between general and specific ability. Specifically, It was found: . . . (1) that the typical musical person has a high grade mentality and shows versatility, particularly in literary and artistic fields; (2) there is a close and definite correlation between musical and mathematical ability; (3) the musical person is likely to have notable linguistic ability; (4) the musical person is likely to show qualities of effective social leadership; (I ) he is emotional, unstable, and not very punctual or scientific; and (6) he is physically healthy and active and endowed with strong neurotic or hysterical tendencies.-1'0 As stated previously, the American studies found little correlation between musical talent and other mental abilities. This seeming contradiction of research results is best explained in terras of Just what is included in the criteria a ° See Chapter Two. 9 James L. Mursell and Mabell Glenn, Psychology of School Music Teaching. NY: Silver Burdett Company, 1938, p. 20. 10 Ibid.. pp. 19-20 7 or musical talent, in each instance. The American research studies are based, for the most part, on the Seashore and Kwalwasser-Pykeme tests of musical talent, which attempt to measure limenal reactions to certain elements of musicality. The German research studies have defined musicality in functional terms, and have stressed the importance of ability to deal with musical materials. This ability has included the ability to appreciate music, whether found in isolation or in combination with executant or creative abilities.'*"*' In commenting on several of the German studies, liursell says: • • . by far the most important conclusion we would draw from the work of Feis, Uaecker and Ziehen, and Koch and MJoen is that distinctive musical ability is a manifestation of a general high level of all-round ability, rathei* than a specific and specialized musical talent that is inherited from the parents.^-2 When musical talent is defined in functional terms, it is not conceived as a pure ability, clearly distinguishable from other more or less isolated, independent, or specific abilities. Musical talent, then, is defined in its relation to general ability, and to certain fairly specific abilities. This study has attempted to learn the relation, if any, of certain musical, mental, and personality traits (in terms of test scores ) to general musicality and certain other specified musical abilities. 11 Ibid.. p. 11 1S Ibid.. p. 18 8 II. SUMMARY STATEMENT INDICATING THE VALIDITY OF THE TECHNIQUE OF SOLUTION It was the purpose of* this study to determine the relationship, if any, between certain mental, personality, and musical factors (as measured by standardized test scores ) and certain musical ability scores derived from rating scales of certain functional musical abilities. Further, it was the purpose of the problem to determine this relation in a practical college situation, when college music students were studied. Since the purpose of this study was of a practical nature, it was appropriate to use, for the collection of a major portion of the data required for the study, such standardized tests as are in general use at the college level. The study utilized: (1) a test of scholastic aptitude, (2 ) a test of reading ability, (3) a test of musical talent, (4) a test of personality, (5) certain ratings of musical abilities, and (6) college applied music grades. The ratings were obtained locally, using a specially constructed rating scale. The test p£ scholastic The American Council on Education Psychological Examination was selected as a source of scores designed to indicate expected college academic success. "The purpose of the American Council on d 9 Education Psychological Examination Is to appraise what has teen called scholastic aptitude or general Intelligence, with special reference to the requirements of most college curricula. The test was found further to be useful and appropriate to the study since it yields the following scores; a linguistic score, a quantitative score, and a total score. These scores are designed, to measure factors known to correlate highly with general intelligence. (For* further discussion of this test and those to follow see Chapter Four, "Sources of Data.*1 Statements concerning the validity, reliability, and other pertinent information will be found in this chapter. ) The test of reading ability. The test battery used in establishing reading ability was the Cooperative Reading Comprehension The Cooperative Reading Comprehension X&gJaL constitute a part of the new cooperative English Test. which is divided into tests of expression and tests of reading comprehension, involving respectively, the active and passive use of the language. • . • The Cooperative Comorehension Reading Tests provide four separate scores; Cl) Vocabulary Score, (2) Speed of Comprehension Score, Co) Level of Comprehension Score, C4) Total Reading Score. 15 American Council Education PgyffiiPlOfiA.gfll fixawAMttjon: Manual o£ Instructions for &1S PgyqhoJLPftAgflJL Princeton, New Jersey: Educational Testing Service, Cooperative Test Service, 1940, p. 2. 14 Cooperative Reading Comprehension Tflgfeft, Inf Concerning Their Construction. Interpretation, and. IfeA. New York: Cooperative Test Service, 1940, p. 1# TJifiL tegt o£ personality. liig, .Fsrg.QnfilitY. ifaygfikPgy15 of Robert G. Bernreuter was used as a source of scores purporting to give some information as to the personality and emotional make-up of the college student. the following scores: The test yields (1) Neurotic Tendency, (2) Self- Sufficiency, (3) Introversion, (4) Dominance, (5) Confidence, and (6) Sociability. This test was standardized with, and specifically designed for, use with college students. The test of musical t.«i en^. The practical aspect of the study was enhanced by the inclusion in the data of test scores derived from the Seashore Measures of Musical £sl££dL* 16 This test battery is widely used in schools and colleges and ittempts to measure acuity of response to certain aspects of musical talent, largely, of an acoustical nature. The 1939 revision of the Seashore measures consists of the following tests: (1) Fitch Discrimination, (2) Loudness Discrimination, (3) Time Discrimination, (4) Timbre Discrimination, (£•) Rhythm, and (6) Tonal Memory. Using the four tests described here, it was possible to make the following comparison: Cl) the relation found between 15 Robert G. Bernreuter, The P?rsnnBn t v Inventory. Stanford, California: Stanford University Press, 1935. ^ Carl E. Seashore, Don Lewis, and Joseph G. Saetvelt, Manual &£. Instruction setiL Interpretations, for tha geagfrgfg Measure Musical (1939 Revision). Educational Department, Radio Corpoi*atlon of America, Camden, New Jersey, 1939. 11 musical ability ratings and these mental and personality test scores with (2) the relation found between the musical ability ratings and the musical talent test scores. The musical ability ratings. In the organization of the study, it became crucial to select a method to obtain, in quantitative terms, a score purporting to measure certain musical abilities as they occurred in the group of college music students studied. The crux of the matter was to define adequately each musical ability for which a measure was desired. The problem was one of developing adequate criteria or evidence of success as applied to musical abilities. In writing on the development of criteria for use in validity studies of vocational tests, Adkins states: Unfortunately there is no universal agreement os to what constitutes vocational success, which is the goal of prediction In the case of civil service tests. What is accepted as evidence is, then, largely a matter of Judgment. Best results are obtained by pooling the Judgments of a number of competent persons as to what variables to include In the criterion. This method of pooled Judgment was used to arrive at a definition for each musical ability Included in the problem. (See Chapter Four for full and explicit discussion of this point. ) Published writings of outstanding educators and psychologists in the field of music* including such names as 17 Dorothy C. Adkins, Construction &£& toftlYPlg S L Achievement Tests. Washington, D. C . : Government Printing Office, 1947, p. 171. IB James L. Mursell, Max uchoen, Carl E. Seashore, and Jacob Kwalwasser, were briefed for definitions of those musical abilities with which the study was concerned. Further, these definitions were submitted to approximately fifty college music teachers for the purpose of arriving at consensus definitions for each of the musical abilities required by the study. These consensus definitions were then used in the study as criteria of those musical abilities. An actual score, purporting to indicate relative status on these musical abilities, was obtained by the use of a rating scale, set up and administered accox*ding to the general procedures outlined by Adkins. 10 Next, the students utilized in the study were rated as to these musical abilities. This rating procedure was carried out by the same faculty grou. who defined these musical abilities. Average ratings were computed for each student, provided an arbitrarily determined standard of reliability was met. These averaged ratings were then included in the study data. Applied music g j In order to have an additional score purporting to measure functional musicality, college grade point averages in applied music were Included in the study data. These grade point averages were derived from grades received in three consecutive terms of applied music 18 Ibid.. p. 232 13 study. The questions to be answered were these: are these personality, mental, and musical talent test scores related to grade point averages in applied music? If so, how do these relations compare in magnitude? Direction of the study. One direction of this study was toward a profile examination of musical talent. To a considerable degree, the study utilizes an atomistic approach to the study of rausicnlity and describes a group of music students in terms of standardized test scores. .Vith these considerations in mind, it appeared that the validity of the technique of solving the problem was enhanced by utilizing several different kinds of tests, each providing several fairly independent scores. After the above variables were obtained, it was possible to carry forward the purpose of the problem, i.e., to determine the relation, if any, of scores received on these personality, mental, and musical talent tests to musical!ty ratings and ratings of certain functional musical abilities of college music students. Briefly stated, the problem purported to determine in terms of test scores, the mental and personality profile of the college music student and to determine the relation of these px*oflle items to functional musical abilities. The study was a statistical one involving wide use of the product-moment correlation. CHAPTER II SUMMARY OF PERTINENT LITERATURE The field of music education has been slow to develop sound psychological tenets upon which to base a psychology of music study. Much energy has been expended in xesearch in music education, the results of whica have in many instances been inconcl nsive. This aura of inconclusion has,in large part, been due to differences in pointof view, or of approach, on the part of many investigators. The following summaries of research studies have been selected to point up two distinct approaches to the study of the nature of musical talent. I. THE AMERICAN STUDIES In the course of the past three decades there has been a great amount of interest in America on the part of many musicians and psychologists in the psychology of musical talent. Consequently, a great many research studies have been concerned with this area. Of considerable importance to this research work have been the published tests of musical talent. The tests most often those of Carl E. Seashore and found inthese studies are of JacobKwalwasser. These music tests have used an atomistic approach to the problem of musical talent, and have utilised selected elements of 15 musical talent, primarily or on acoustical nature. underlying assumptions of this approach arcs The Cl) that musical talent is a composite of many separate abilities and (2 ) that some of these abilities can be isolated from musical contexts and measured specifically. position: Xlursell states, concerning this "It is fair to say that they of Musical Talent the Seashore Measures represent an attempt to apply to the measux*ement and diagnosis of musical capacity the sensationalistic position in psychology."**" As later discussion will point out, this approach has been something of a negative one, since research has failed to show that those individuals scoring high on the Seashore and Kwalwasser-Dykeraa tests will also prove to be musical. Research has pointed out, however, that those individuals scoring high on a criterion of musicality also score high on certain of the tests of the Seashore and Kwalwasser-Dykema test batteries. V a l i d a t i o n studies. Larson, p in studying the relation between musical talent and the ability groupings in a well organized high school music program, found that although the 1 James L. Mursell, The Psychology of Music. iff. ;v. Norton, 1937, pp. 299-300. New Yorks 2 William G. Larson, Measurement of Musical Talent for the rredlctlon of Success in Instrumental Music. Psychological Monographs, No. 181, University of Iowa Studies in Psychology. Iowa City, Iowa: University of Iowa Press, 40: 33-73, 1930. 16 beginning instrumental classes had musical talent averages approximating that or an unselected group, the most advanced orchestra group had averages that must admittedly belong to a selected group. In terms of Seashore percentiles, the beginning group had a total score averaging 52.1, while the advanced orchestra group had total scores averaging 73.2. Larson states; Since the members of the high school advanced orchestra are selected on the basis of their ability to perform, it is reasonable to conclude, judging from the averages of the beginning instrumental classes, that the capacities of musical talent as measured by the Seashore tests, have been an influence in the selection of the membership of this orchestra; and that groups at Intermediate stages of advancement have been correspondingly affected.3 Relative to the effect of training on these scores, Larson points out that; . . . the above interpretations rest upon the assumption that these measures are elemental in that they are not affected to any great extent by training. This conclusion rests upon experimental facts derived over a period of the last twenty-five years, the cumulative results of which have largely been responsible for furnishing us with a psychology of music.4 5 Gilbert made a somewhat similar study in which he attempted directly to show the relationships between musical 5 Ibid.. p. 55. 4 Ibid., p. 62. 5 J. R. Gilbert, "The Traits of Secondary School Instrumentalists and Their Relationship to Achievement in Instrumental Music." Unpub. 14.A. Thesis, Syracuse Univ., 1943. talent and achievement. In working with one thousand boys and girls of high school age using the Kwalwasser-IJykema Tests, Gilbert found a correlation of .76 for the entire group with teachers' ratings of achievement. Correlations broken down in terms of length of study, follow: one year study, .69; two years, .68; three years, .73; four years, .86; five years, •77; six years, .75. These ratings were arrived at by the use of a linear scale ranging from one to ten. These studies point up the fact that those who succeed in music possess certain elemental capacities of an acoustical nature. This is not the same as saying that those possessing these elemental acoustical cax^acities will succeed in music. The next study to be reviewed offers a point of view which is pertinent to a clear understanding of the meaning of the correlation coefficients found In the Gilbert study. Stanton, in an ambitious validation study covering a ten year period used the Seashore Tests in combination with the Iowa Comprehension Isfii. (a group test of general intelli­ gence). The Seashore Measures for pitch, time, consonance, intensity, and tonal memory, plus the intelligence test, were administered to all entering students. On the basis of this battery, individuals were segregated into five classes as 6 liazel Stanton, Measurement Eastman Experiment. Iowa City, Iowa: Tress, 1935, 140 pp. Ihfi. University of Iowa 18 follows: discouraged, doubtful, possible, probable, and safe. Their later achievement in the conservatory was studied. Annual survival, avoidance of dismissal, attainment of scholar­ ships and honors, recital appearances, and graduation were the most important factors considered. In all of these respects an increasing degree of success was demonstrated in passing from the low to the high groupings. The most typical and directly convincing results were those for graduation. Of the discourage! croup, 17 per cent graduated; of the doubtful group, 23 per cent; of the possible group, 33 per cent; of the probable group, 42 per cent; of the safe group, 60 per cent. The students were not Informed as to their classification on the tests, so that a low grouping had no uni'avorable influence, and the whole machinery of measuring educational success was isolated from any Influence by the test classifications. Mursell says regarding this experiment: These findings are undoubtedly significant and of practical value. The wording of the various classifi­ cations is open to some objections, for it would seem inappropriate to call individuals whose chance of graduation is 60 per cent ‘safe', and a 42 per cent chance of success Is not what one ordinarily means by the term •probable*• But the point is not of major importance, and it is clear that the battery possesses considerable predictive value. But we cannot regard the results as in any way an adequate validation of the Seashore Measures of Musical Talent. It should be clearly understood that the groupings were fox*med not on the Seashore Tests alone, but on those tests combined with an intelligence test.7 7 James L. Mursell, The Psychology g£. Music. Vi, Vi, Norton and Company, 1937, pp. 298-299. NY: Kursell concludes that the results of this experiment *. . . furnish no proof whatever that the Seashore tests given Independently of any other measures will yield a valid Index o of musical capacity.n Musical 3J& .MSJ&JaJL l£rifcJL- There have been many studies of this nature, usually dealing with correlations between scores earned on either the Ceashore or KwalwasserDykema music talent tests and personality and intelligence tests. Gilpin, 9 in working with Junior and senior high school students, attempted to find the relation between musical talent and certain mental and personality aspects. From two schools, 25L students in grades seven through twelve were tested. He concluded that musical talent and intelligence are not closely related. This is born out by many other studies of a similar nature, as we shall see. The corre­ lations typically found are positive, but low, usually not above .35. Gilpin also found that musical talent and social and emotional adjustment are not closely related. Here again the correlation coefficient was positive, but low, .378. 8 Ibid.. p. 299. 9 G. Noble Gilpin, "A Study Correlating Scores on Kwalwasser-Dykema Tests of Musical Talent, Washbume Thasplc Personality Inventories, Otis Quick Scoring Mental Ability Tests, and an Original Phonophotographic Test.” Unpub. &. A. Thesis, Syracuse University, 1941. VJenaas,1® in n study having many’ aspects in common with that or Gilpin, reported above, round correlations between Kwalwasser-Dykema test scores and intelligence test scores rangi;ig f*rom *714 to .211 and tending to decrease with age. Between Kwalwasser-Dykema scores and personality ractors, VVenaas round positive, but low non-significant correlations. The personality Tactors correlated were those or the Case inventory by J. B. Mailer and included controlled association, personal and social adjustment, honesty, and ethical Judgment. The Otis and Kuhlmann-Anderson tests of* intelligence were used Groups studied were taken rroin grades six through twelve. In another study involving the use or the Kwalwasserlykema Tests s£_ Musical Talent. Wagner11 round a correlation or .402 between musical talent and well adjusted personalities v.agner points out that all correlations between subtests (Kwalwasser-Dykema Tests or Talent and »Vashburne Social Adjustment Inventory) Tavored the musically talented student• Sigurd B. Wenaas, "A Study or the Aelatlon Between Musical Ability and Various Intelligence, Scholastic, and Personality Factors." Unpublished M. A. Thesis, University or Idaho, 1940. ^ Doris Wagner, "A Comparison or Earned Scores by Junior high School Pupils in the 7/ashburne Social Adjustment Inventory and the Kwalwasser-Dykema Tests or Musical Talent." Unpublished M. A. Thesis, Syracuse University, 1946. SI In an investigation in which the Seashore tests were used, Immel12 found no appreciable relationship between emotional status and musical capacity among high school pupils. Measures of emotional status were obtained by use of the Bernreuter Personality Inventory, and the Allport Ascendance" Submission Teg£. and the Thurstone ffgEpom&Afcy. &£&£&&&• Musical performance and pant*! traits. Gilbert, 13 in using teachers' ratings for a measure of performance achieve­ ment, found correlations between performance achievement and Kwalv;asser-Dykema scores ranging from -761 for the entire group Cone thousand) studied, to -866 for a group of 123 students selected on the basis of four years study. For the entire group of one thousand, who had had musical training, Gilbert also found a correlation of -574 between the Otis Intelligence Test scores and the Kwalwasser-Dykema test scores14 Lamp, in a study in which he used an exposure course (a forty period instrumental trial), found that in comparison 12 Earle Barnard Immel, "An Experimental Investigation of the Relationship Between Musical Capacity and Emotional Status of High School Seniors.” Unpublished M. A. Thesis, University of Southern California, 193913 J. Richard Gilbert, ”The Traits of Secondary School Instrumentalists and Their Relationship to Achievement in Instrumental Music,” Unpublished M. A, Thesis, Syracuse University, 1943. ^ Charles Lamp, "The Determination of Aptitude for Specific Musical Instruments-” Unpublished Ph. D. Thesis, University of California, 1935- 82 to the predictive values or his aptitude test (a perf ormance test given at the termination of the forty period exposure)* the mental and physical tests* (Seashore Measures of Talent, and certain physical measurements* such as finger taper for violin* also used in the study as a basis for comparison )* offer no valid prediction to serve as a basis for the selection of instruments most suitable for individual students* II. THE GERMAN STUDIES Contrasting the work of the American research workers is the work of the German researchers. Here* the approach to the problem of musical talent has been, for the most part, that of the "gestalt" psychologist using musical criteria as a basis for prediction. Here the underlying assumption is: musical talent is expressed by the whole personality and therefore cannot be ascertained by dividing the musical personality into fragments to be measured individually, but must be measured as a unit functioning in a unified situation. RaJLflfrlafl qL mfels.aJL.itaL £& ath ay lypep o£ & £ lll£ z > Mursell has written an excellent summary of several German studies of particular pertinence. First of all we have the work of the Pannenborgs 15 H. Musikers." J. and W. A. Pannenborg* "Die Psychologle des Zeltschrift fur Psychologic. 73: 91-136* 1915. 23 16 and or filler. The .former, In their study* investigated the abilities of 423 musical adults, 21 composers with whom they used the biographical method, and 2757 school children between the ages of 12 and Id. It should be stated that they found a high measure of agreement between their three groups, so that we may conclude that musicality has about the same psychic and cultural characteristics wherever found* The latter (Miller) studied the school records of students in a teacher training Institution where music was required, men only being investigated* The characteristics of the musical personality as revealed by these studies are as follows: (a) the typical musical person has a high grade mentality and shows much versatility, particularly In literary and artistic fields* (b) There is a close and definite correlation between musical and mathematical ability* (c) The musical person is likely to have notable linguistic ability* The work of Peis1” emphatically confirms this finding* (d) The musical person Is likely to show qualities of effective social leadership. ( e ) He is emotional, unstable, and not very punctual or scientific, (f) lie is physically healthy and active and endowed with strong neurotic .or hysterical tendencies. The general picture is that of a high grade, nervously organized and high-strung personality, urgently needing free and varied outlets for personal expression, and capable of great contributions, though these need not be kept in the focus of music education* Another point to be considered here is the relation­ ship of musical ability to general intelligence* And here we find a sharp cleavage between the German and American studies. Fels, Revecz,13 the Pannenborgs, Miller, and others are unanimous in finding that musicality and high intelligence go together. Seashore, on the otnar Richard Miller, "Uber Musikallsche Begabung und lhre Bezlehungen zu sonstigen Anlagen*M Zeitschrift £u£. PpyghfflgftAft* 97: 191-214, 1925. 17 Oswald Fels, Studlen uber die Qenealogie yfid Pffy.cho.-:. logie der Musiker. iVeistoaden: J. F. Bergman, 1910, 97 pp. 16 Geza Revecz, The Psychology New York: Harcourt, Brace and Company, 1925, 180 pp. 24 hand, finds comparatively little relationship between these two functions. Hollingworth,19 again, using the Seashore Tests, finds no special musical sensitivity in children testing above 135 I. Q. which means of course, a group of very high intelligence. How can we explain these seemingly flat contradictions? The answer is not difficult, the American work is all based upon the very special Seashore Tests, while the German work, as is clear from our description of it given above, takes a much broader, more functional criteria of musicality. There may be no relationship between Seashore Test performance and intelligence and still be a very close one between functional musical ability and intelligence. As a matter of fact, this is precisely our own opinion. We feel that there is ample Justification for stating that musicality goes with high general intelligence.20 Relafripji o£ HH £l3. i® ► ciT - OS osr 09 08 4 ~^o -aeqactf 34 determine readiness Tor college music study In applied music. One student In the group was allowed to enroll but was not Judged ready to earn college applied music credit. All or these students were required to take two years work In theory. This theory work is designed (1) to acquaint the student with the materials or music, and (2 ) to encourage the student's responsiveness to music In a variety or activities Final grades achieved by the group in the rirst year's work In theory were distributed as rollows: 21.93 per cent received the grade or HA," 39.37 per cent received the grade or "B,H 30.96 per cent received the grade or "C," 3.67 per cent recelvec the grade or "D, H and 3.67 per cent received the grade aT "F." The grade or "A" is the highest passing grade; the grade or "D" is the lowest passing grade. grade. The grade or “F" is a railing This distribution or rirst year theory grades is shown on Qraph 4.. (For rurther evidence or the musicality or this group see Chapter Five, Summary or Study Data. InTormatlon concerning other music grades and musical abilities will be round in that chapter.) II. CONDITIONS OF DATA COLLECTION As part or the Michigan State College orientation program, each entering student is required to take several standardised group tests. The group oT students used in this study took Applied 9 ^-Instrumental _gihgsnecftl w.JhffiS3L Science & Arta t t • I i * ................ f -* * Music /Therapy n Ntunber cases Graph $ hisraiBifc i o n !o f "s MAJOR UfSfePR •GROUEXHGS i ' | NTS BY 1 r L-.J 34 determine readiness Tor college music study In applied music* One student In the group was allowed to enroll but was not Judged ready to earn college applied music credit* All or these students were required to take two years work In theory* This theory work Is designed (1) to acquaint the student with the materials or muslca and (2 ) to encourage the student*s responsiveness to music In a variety or activities Final grades achieved by the group in the rirst year*s work in theory were distributed as rollows: 21*93 per cent received the grade or "A,* 39*37 per cent received the grade oT "B," 30.96 per cent received the grade or "C," 3.87 per cent received the grade or "D," and 3.87 per cent received the grade oT "F." The grade or "A" is the highest passing grade; the grade or **D* is the lowest passing grade* grade. The grade or "F" is a Tailing This distribution oT Tirst year theory grades is shown on Graph 4.(For Turther evidence or the muslcallty oT this group see Chapter Five, Summary oT Study Data* ZnTormatlon concerning other music grades and musical abilities will be round in that chapter*) II. CONDITIONS OF DATA COLLECTION As part or the Michigan State College orientation program, each entering student Is required to take several standardized group tests* The group of students used In this study took 60 50 40 50 80 10 ; Graph 4 DISTRIBUTION OF THEORY (GRADES 36 such tests either as entering freshmen ox* as entering transfer students. In most Instances, such tests are administered prior to ox* during the first term of the student*s registration. A second phase of the college orientation program consists of tests, of a less general nature than those referx»ed to above, given by various academic departments in the college. The examiners of the Michigan State College Music Department administered a musical talent test to this group of students. The College Counseling Center administered an additional test to this group of students for the specific purposes required by this study. This was a test of personality. Each testixig situation referred to above was conducted by experienced examiners. Results were machine scored. The study required ratings of certain musical abilities as part of the data. These ratings were made by 51 members of the Michigan State College music instructional staff. The ratings were based on the professional Judgment of this faculty group, and were secured by the use of a carefully planned rating procedure. III. (See Chapter Four. ) SUMMARY It appears that the group of students utilized in the study is a musical group of students. There Is evidence that these students as a group were interested In music several years px’ior to entering college. Some individuals 37 began music study as early as rive or six years or age. There is evidence that this group or students had ability in music beyond an elementary level prior to college enrollment end that the group as a whole had suTricient ability to do acceptable college work in music. It also appears that this group or students v;as especially suitable to the study since it was possible to utilise these stuaeats in a large variety oi* situations for the purpose or obtaining adequate data. At the same time, by utilizing this group It Aas possible to gather all data under very Tavorable conditions. CHAPTER IV SOURCES OF DATA The data collected Tor use In this study were obtained through the following sources: (1 ) Use of the f oilowing published standardized psychological tests; (2 ) (3) (a) The American Council on Education Psychological EfcflP&ftftfckfla* College level. Cooperative Test Division, Educational Testing Service, Princeton, Hew Jersey. (b ) The Cooperative ghehglffiR IsBXJL* Cooperative Test Service, 15 Amsterdam Avenue, Hew York, Hew York (c) The Bernreuter Personality Inventory. Robert 0. Bernreuter, Stanford University Press, Stanford University, California, 1935. (d) The Seashore Measures of 1939 Revision, Camden, New Jersey: Victor Division, 1939. Ratings of musical abilities (a) Muslcality. (b) Ability to sight-read. (c) Ability to perform. Grade point averages In applied music. RCA * 39 I. THE AMERICAN COUNCIL ON EDUCATION PSYCHOLOGICAL EXAMINATION According to the manual or Instructions: The purpose of the American Council on Education Examination Is to appraise what has been called scholastic aptitude or general Intelligence, with special reference to the requirements of most college curricula. A large number of different tests have been used for this purpose. It has been found that, in general, linguistic tests give higher correlations with scholarship In the liberal arts colleges than do quantitative tests. This higher correlation is probably, in part, due to the fact that most of the freshman courses in the liberal arts colleges depend more upon linguistic abilities than upon the abilities involved in quantitative thinking. For the scientific and technical curricula the quantitative tests may be more significant. The test forms should be found useful in handling those problems in which it is advisable to distinguish a student’s mental abilities from his high-school preparation and his industry. Faculty action in the case of a student who is failing can be intelligently guided if one has some means of knowing to what extent the student has applied himself to his college work, to what extent his high-school training meets the requirements of his college course, and what his mental abilities are. Very different faculty action can be taken, depending on which of these three factors may be held primarily responsible for a student’s failure. It is to be hoped that these psychological test forms may lead to the early discovery of bright students. In those colleges where sectioning of classes In accordance with ability or preparation is customary, these test forms may serve as part of the evidence upon which the sectioning is based. There is one form of the current edition. All students take the test in exactly the same way, marking their answers on separate answer sheets. The answer sheets are scored by hand or by means of the electric scoring machine. Since the task x or the subjects is identical regardless of the way in which the test papers are scored, only one set of norms is required. The same test booklets can be used for several grcips of students provided that the students do not make marks in the booklets. The examination consists of the six tests that have been used for several years. The order of the tests has 40 been arranged to alternate linguistic and quantitative tests because of the fatigue element. All of the test Items have been Included In several test experiments with factorial analyses to determine the primary mental abilities. These studies have justified the grouping of the six tests In two general classes as follows: Quantitative Tests: (Q-score) Arithmetical Reasoning Number Series Figure Analogies Linguistic Tests: Same-Opposite Completion Verbal Analogies (L-score) It Is not recommended that the six separate test scores be used for any counseling, but there seems to be Justification for using the two principal subscores as well as the total or gross score in this manner.1 Hereafter, in this study these Psychological Examination scores will be referred to as ACE-L, ACE-Q, and AGE-T. ACE will refer to the test, in general. A survey of research concerned with the various editions of the ACE revealed much interest on the part of researchers in (1 ) prediction of academic success and (2 ) the relationship of intelligence, as measured by this test, to academic success. The following statement of research done with the various editions of the ACE, is based on summaries of some thirty selected studies reported since 1937. Findings that appear to be typical, and in general agreement, have been selected from several studies for quotation here. Hauser states: The value of Intelligence tests for predicting college success has been widely Investigated. Llepold, American Council on Education Psychological Examlnation: Manual of Instructions,_1949 edition. Princeton, Ni J . : Educational Testing Service, Cooperative Test Service, p. 2. 41 Bovee and Froehlich, Bernreuter and Goodmanf Stult and iiudson, Bryan, Mitchell and others conclude from their research In various subject matter areas with different tests, that Intelligence rating alone does not exert a decisive Influence on scholastic success. . . . That Intelligence Is never the sole factor in any real life situation Is found in the researches of Burt, Gates, Miller, Madsen, Shewman, Binet and Glenn and many others. Freemen, summing up the prognostic values of intelligence tests, says that most Intelligence tests are useful In predicting educational achievement but they are not sufficient in themselves.2 In the light of these statements, it is not surprising to find correlations between ACE scores and grade point averages in various subjects occasionally as high as .67 (between "L" scores and so-called •'verbal** subjects ), and ranging as low as .19 (between "L" scores and so-called "quantitative” subjects). Schmitz, in a prediction study of the relationship of vari ..us test scores and college grade quotients concludes: "The ACE ranks second (to high school quotient) as a criterion of success, having a correlation of .503." The following coefficients are reported: College quotient: with with with with high schoolquotient ACE Armv Alpha Purdue Placement Isfll .644 .583 .576 -463 2 Luellen J. Munn Hauser, "A Comparative Study of the Intelligence of University Freshmen Enrolled in Business and Liberal Art Schools •" Journal of Educft*rionai Research. 43: 49, 1949. 3 Sylvester B. Schmitz, "Predicting Success in College." Journal of Educational Psychology. 28: 466, 1937. 42 Schmitz points out that the ACE ". . . Is only slightly better than the Army Alpha as a criterion or success In college."4 Lanigan^ reports the Tollowing correlations between selected subject matter areas and ACE-T scores: English Social Studies Language Mathematics Science Fine Arts *891 .483 .830 .837 .531 .381 ACE .385 .501 .828 .384 .442 .364 Minn. .548 .393 .483 .194 .458 .317 In reporting critical ratios of the differences between the high and low achievers in each of the six subject matters listed above. Lanigan concludes: "These findings Indicate the American Council on Education F s v c h o J l c * n Examination is a more usable instrument for predicting a critical score above which groups succeed and below which many tend to fall ox- to receive low marks. Somewhat less optimistic findings are repox-ted by Wallace in a study presenting correlation coefficients obtained between the 18 largest and most usual courses of the fix-st semester 4 Ibid.. p. 466. ® Mary A. Lanigan. "The Effectiveness of the Otis, the ACE and the Minnesota subject Speed of Heading Tests for Predicting Success in College." Journal of Educational Research, 41: 290-296, 1947. 6 Ibid.. p. 293. 43 and quantitative, Linguistic, and Total scores made by freshmen in the fall of 1947 at the University of Michigan. Wallace summarizes these findings as follows: • • • (1) all correlations were small* (2 ) The highest reiationsaip between test scores was a multiple •49 between English and the combined Quantitative and Linguistic scores. (3) Means and sigmas for the Quantitative and Linguistic scores show little differentiation.7 Studies of vftllriltv and reliability* In a study of the validity and reliability of the 1938 edition of the Psychological nation. Seder reports the following reliability coefficients for the Total score, .952; for the Quantitative score, *866; and for the Linguistic score, *953* Seder concludes: The 1938 edition of the American Council on Education Psvcho.io~--lr»gil Examination, although it has been changed considerably, seems to be similar to earlier editions as far as the total score is concerned. The total score continues to be internally consistent and is highly related to total scores on the 1937 edition of the examination. The Linguistic scores tend to be more highly correlated with measured achievement in English, for language, history, and science than the Quantitative scores; the reverse is true of correlations in the field of mathematics.s L. Wallace, "Differential Predictive Value of the American Council on Education Psychological Examination." Cchool and Society, 70: 24, 1948. 8 Margaret Seder, "The Reliability and Validity of the American Council on Education Psychological Examination." Journal of Educational Research, 34: 101, 1940. 44 Votaw 9 reports a correlation between the ACE-T and the Otis Intelligence test or .76. Lanigan^ reports a correlation between these tests or *652, and between the .Minnesota Speed or Reading Test and ACE-T, reports a coeTiicient oT .334. A studv* included In the Wilson College Studies in Psychology makes the rollowlng comparative statement: The correlation or the American Council on Education tests, Revised StanTord-Binet test and the W'echslerBellevue verbal scale with grade point averages are all., approximately equal (correlations or about .50 to .55). **■ Summary. These Tindings or research (reported above) done with the ACE tend to be typical and tend to support the rollowlng statements: (1) The ACE-T scores tend to predict college scholastic success about as well as, or a little better than other paper and pencil tests oi* intelligence. This relation between ACE-T scores and college marks can probably be indicated by a product-moment coerriclent or .45 or better. ^ David F. Votaw, "Regression Lines ror Estimating Intelligence Quotients and American Council on Education Examination Scores." Journai oT Educational Psychology. 37: 179-181, 1946. 10 Lanigan, o p . clt. ^ Wilson Studies in Psychology, "A Comparison or the ■Vechler-Bellevue, Revised StanTord-Binet, and American Council on Education Tests at the College Level." Journal or Psychology. 14: 325, 1942. 45 (£) The .reliability or the ACE scores is acceptable for most uses, being somewhat better In the Instances of the Linguistic and Total scores than in the instance of the Quantitative score. (3) Research findings have not agreed as to the use­ fulness of the Quantitative and Linguistic scores for differential prediction of college success. Differing approaches to the problem, and differing points of view of investigators, have tended to make for inconclusive findings II. THE COOPERATIVE READING COMPREHENSION TESTS According to the sheet of information concerning the construction, interpretation, and use of these tests: The Coop.sr&t.lvg. QPP.P.f.^hjg.n^lpji X.gff.tjjg constitute a part of the new Cooperative English Test. which is divided into tests of expression and tests of reading comprehension, involving respectively, the active and passive use of the language. . . . The Cooperative Reading Comprehension Tests provide four separate scores: Cl) Vocabulary Score, (£ ) Speed of Comprehension Score, (3) Level of Comprehension Score, (4) Total Reading Score. The reading sections of these tests are based on the belief that reading comprehension is essentially a thinking process, a process which requires mental facility in manipulating verbal concepts, a background of experience, and skill in the mechanics of reading, determine the level of comprehension which an individual may attain and also the maximum speed with which he is able to read and comprehend materials of a given level of difficulty for a specific purpose. Almost all previous reading comprehension tests have measured a combination of vocabulary level, speed of reading, mechanics of reading, and ability to answer questions based on the facts stated in certain passages. Other 46 components or the ability to read with understanding have largely been neglected. In the Cooperative Comprehension Testa, an effort has been made to Include these neglected factors by emphasizing the measurement of the thinking processes in reading, the importance of which has i*ecently been emphasized anew by Investigations in the field of semantics. The validity of a renuing test is determined by the extent to which it measures the skills actually Involved in the reaalng process. To establish the validity of the goopfijrgtlyQ, Reading Comprehension Zfifi&S. a thorough analysis of the reading process was made. On tne basis of this analysis, the outline for the tests, . . . was developed. Items were then constructed to measure the skills Included in the outline. As stated above, the test Is designed to measure the thinking processes in reading on the grounds that effective reading Is not a mechanical process but an active reasoning and assoclational process. Validity and difficulty indices were obtained for each item in the experimental forms of these tests. Items at the ^roper levels of difficulty having relatively high correlations with the total scores were then selected for inclusion in the final forms, revisions being made on the basis of tne Item analysis. In the vocabulary sections of botn Higher and Lower Level tests the average correlation coefficient between the individual items end the total score is .52. For the reading section of the Lower Level the corresponding value is .45, and for the reading section of the Higher Level It is .40. Because the more dis­ criminative items in the reading sections tend to have been placed first in the tests, those items which actually determine an individual's score tend to have an average validity index greater than the average values presented above. This is a desirable feature of the tests which operates to increase their accuracy of measurement. A study of the lntercorrel&tlons between the three part scores of the Cooperative Reading Cpppy^^nglPft Tggtg indicates that they measure closely related abilities. An effort was made to reduce the effect of word knowledge on the comprehension scores by controlling the vocabulary level of the reading sections. That this effort was fairly successful is indicated by the only moderately high correlations between Vocabulary and Speed of Comprehension (.75 at both Lower and Higher Levels). The correlations 47 between Level or Comprehension and Speed of Comprehension are *89 at the Lower Level and .87 at the Higher Level* The V o c a b u l a r y Score Indicates the extensiveness of the individual's word knowledge* The time limit Tor this section Is long enough so that, except Tor a Tew Individuals whose mechanics of reading are extraox*diaarily poor, speed of word recognition plays little part In determining the Vocabulary Score* The Speed of Comprehension Score represents the product of the rate at which an individual has attempted to comprehend the test material and his success in comprehending it* It is not, like many speed of reading scores, merely a measure of the number of words read without regard to the thought content* The Level of Comprehension Score provides a measure of the ability of the student to comprehend materials of increasing difficulty at the rate at which he chooses to work. It is a measure of ’•power" or "depth" of compre­ hension, indicating the extent to which a pupil is able to grasp the full import of what he reads* The Total Reading Gcore is a composite score in which each of tae other tJiree scores has equal weight* It may be regardeu as an measure of linguistic a b i l i t y and should prove to be an excellent index of scholastic a p t i t u d e * 1 2 Hereafter, in this study the Coo iterative Reading Comprehension Tests will be referred to as CRCT-V, CRCT-R, CKCT-C, and CRCT-T. CRCT will be used to designate this group of tests in general* 12 Cooperative Reading Comprehension Concerning ThSlf Kew York: Cooperative Test Service, 1940* IflCPffflMAqa 48 III- THE BERNREUTER PERSONALITY INVENTORY The Bernreuter Personality Inventory is a paper and pencil test or personality. It is a multi-trait test and H. • • was constructed to determine the feasibility of estimating more than a single personality tx>alt at a time* It assumes that the Integrated behavior of an individual in any situation view • • .H13 m ay be Interpreted from various points of The test uses a test sheet and an answer sheet which can be machine scored. Bernreuter scores for this study were scored by machine* The test is intended for use with high school, college, and adult populations ana provides norms for these groups, Loth male and female- Raw ecores are converted to percentile rankings accox’ding to a conversion table* The test consists of 125 questions to which the subject may answer Yes. No. or Credit for each trait is given accox*ding to Bernreuter's scoring system* Flanagan14 has also provided the Bernreuter test with additional scoring keys. These several scales are described by Bernx-euter as follow High Bl-N. The individual wno scores high on the Bl-N scale shows a tendency toward a neurotic condition* 13 Robert O- Bernreuter, "Validity of the Personality Inventory," Personnel Jmipwi. 1 1 : 383, 1933* ^-4 Irving Lorge, "Personality Traits by Fiat: A Correction," Journal of rmAl PsvchQlogv‘ 26 * 654* 1935. 49 Such an individual often feels miserable# is sensitive to blame, and is troubled by useless thoughts, by shyness, and by feelings of inferiority. He feels shut off from other people, he frequently day-dreams, and worries both over things that have happened and over things that may happen. The individual who scores low on the Bl-N scale is an emotionally stable person. He is rarely troubled by moods, by worries, or by the criticisms of others. He is self-confldent, and is a doer rather than a daydreamer. High Bg-S. The individual who scores high on the B2—S scale is a self-sufflclent person. He is able to be contented when by himself. He prefers to work alone and depends upon his Judgment in reaching decisions and in formulating plans. Low B2-S. The individual who scores low on the B8-S scale is dependent upon others for his enjoyments. He likes to be with other people a great deal, and prefers company both while working and during leisure hours. He prefers to talk problems over with others and to receive advice before reaching decisions. High B3-I. The individual who scores high on the B3-I scale is introverted in the sense that he is introspective and is given to autistic thinking. He shows the symptoms of a neurotic condition which are typical of those individuals who score high on the Bl-N scale. Low B3-I. The individual who scores low on the B3-I scale is extroverted in the sense that he rarely substitutes day-dreaming for action. He is emotionally stable and possesses the characteristics of those individuals who score low on the Bl-N scale. High B4-D. The individual who scores nigh on the B4-0 scale is dominating in face-to-face situations with his equals. He is self-confident and aggressive, and readily assumes a position in the foreground at social functions. He converses readily with strangers or with prominent people and suffers no feelings of inferiority when doing so. Low B4-D. The individual who scores low on the B4-D scale is submissive in face-to-face situations 50 with his equals. lie lacks self-confidence, keeps in the background at social functions, and rarely takes the initiative in directing people or activities, ile experiences feelings of inferiority and is reluctant to meet important personages•16 The Flanagan scores are described by Bernreuter as follows: A measure of confidence in oneself. Persons scoring high on this scale tend to be hamperlngly selfconscious and to have feelings of inferiority; those scoring above the ninety-eighth percentile would probably benefit from psychiatric and medical advice. Those scoring low tend to be wholesomely self-confident and to be very well adjusted to their environment. JES.-&. A measure of sociability. Persons scoring high on this scale tend to be non-social, solitary, or independent. Those scoring low tend to be sociable and gregarious.16 A summary of research done with this test reveals a great amount of interest on the part of researchers in this test and in the area of personality testing in general. This summary of research done with the Bernreuter Pf>rsr>^plltv Inventory has involved some 140 research studies Including principally Cl) investigations into the reliability and validity of the inventory, and (2 ) studies of correlation with many different variables. 15 Robert G. Bernreuter, "The Theory and Construction of the Personality Inventory,** Journal of. Social Psychology. 4: 402-403, 1933. 16 David G. Ryans, "A Tentative Statement of the Relation of Persistence Test Scores to Certain Personality Traits as Measured by the Bernreuter Inventory," Pedagogical Seminary and Journal of Genetic fiftyqjifllffgy* 54: 230, 1939. 51 In a study of the Bernreuter Person**? it.v Inventory as a measure of student adjustment, Stogdill and Thomas state: The test apparently finds its most useful application at the college level, rather than with younger people. It Is more adequate with the college group than with the definitely psychotic or neurotic inmates of state institutions• It appears to be more useful In the determination of the Introverted and submissive Individual than with the opposite types. The significance of a low score on the scale for measuring neurotic tendency (Bl-N) is not well established. Scores in the middle range on each of the six scales of the test seem to be highly relatea to tne possession of a desirable pex-sonality as Judged by criteria. As a measure of adjustment the Bernreuter Personalitv Inventory appears to be very helpful In discriminating between well-adjusted and maladjusted students.1 ' In a study similar to the one cited above, Fisher and Hayes conclude: There is a significant and reliable relation between high scores on the Bernreuter Personality Inventorv (taken on entrance to college ) and serious maladjustments un­ covered later in college. Scales F2-B, Bl-N and B2-S are most important in this connection. Scoring of the other scales adds to the effectiveness of the test in predicting possible maladjustments only when they are used to corroborate a high score on one of these three.13 In an imposing summary of 147 studies concerned with research on the Bernreuter P e r s o n a l i t y Inventory. Super 17 Emily Stogdill and Minnie E. Tno;r:as, "The Bernreuter Pex-sonality Inventory as a Measure of Ltuuent Adjustment,** Journal of Social Psychology. 9: 313, 1938. 18 Willis Fisher and Samuel P. Hayes, Jr., "Maladjustment in College Predicted by Bernreuter Inventory Scores and Family Position,** Journal of Applied Psychology. £5: 96, 1941. 52 makes the rollowlng cautious statement: This brleT synthesis or Tlndlngs concerning the Bernreuter Pftypoflpiflfry Inventory points to the conclusion that It has considerable validity as a research Instrument; that when properly used It has some value In work with Individuals; that In either type or work care must be taken to secure adequate rapport; and that, some situations may be such as to make such rapport out or the question. In a validity study, St. Clair and Seegers summarize as rollows: Examination or FI and F2 scores (measure or conTidence and sociability) or certain students, whose responses to a questionnaire might lead one to expect certain abnormal personality traits, Indicates that the FI 8core possesses a degree or validity as a measure or selT-conTidence. The FI and B1 scores seem to measure nearly identical traits. However, certain Inconsistencies were apparent when the F2 scores were analyzed. A multi-modal distribution oT the F2 scores oT students who were selected Tor membership In Traternltles and sororities and scores oT a number oT student leaders conrirmed the impression that a high F2 score is not consistent)-/ an Indication oT non-sociability. Examination and analysis or the Bernreuter scores oT the same Individuals Indicate Tairly deTinite proTiles which seem to depict personality types. These proTiles ore determined by studying the Bl, B2 and B4 scores (measures oT neurotic tendency, selT-surriclency and dominance) In their Inter­ relationships and the B2 score seems especially Important. Two proTiles have been delineated tentatively, ProTlle I, It appears might be associated with a withdrawal tendency. Profile II, probably Is Indicative oT leadership. In general, the evidence presented strongly supports the view that the proTlle approach to the Interpretation oT the Bernreuter scores presents a TruitTul Tleld Tor research ^ Donald E. Super, "A Review oT Research,** £ail$BfiUL St. Psychology. 9: 120, 1940. 20 Walter F. St. Clair and J. Conrad Seegers, "Certain Aspects oT the F scores or the Bernreuter Personality Inventory," Journal s L 29: 1938. 53 Lorge, In an investigation into certain statistical considerations relating to the BPI (both Bernreuter and Flanagan keys) states: "Flanagan has provided the Bernreuter Perftfffifliity Inventory with a set or independent keys which results also in consistent scores." 0 1 In a study of the validity and reliability or his Inventory. Bernreuter states as follows: Twenty-foui* determinations or the split-half reliability or the rour scales or the Personality Inventorv averaged .86, the emotional stability and dominance-submission scales showing the highest reliability. The validity of the scales In predicting scores on Thurstone's Neurotic Tendency* Allport’s A-S Reaction Study, Laird’s C2 Intro-Extroversion Test and the author's Self-Cufficiency test is very high* the lowest coerricient being a .84 and the most rrequently round values appi'oximating 1.00. Correlations ranging between .56 and .67 were obtained with seir-ratings or admittedly low reliability.*-2 In an exhaustive sux-vey oi‘ research done on the Bex*nreuter P e r s o n a l i t y Inventory. Patterson makes the following statements concerning its reliability: The x*e.Llability of the Inventory is uniformly high, or relatively high, from study to study. Averaging the split-half and odd-even coefficients reported in half a dozen studies Indicates that .85 may be taken as the best measure of the i%eliability of most of the scales. B2-S and F2-S appear to be rather consistently slightly less reliable, perhaps .80 rather than .85. Test-retest 2*- Irving Lorge, "Personality Traits by Fiat: A Correction," Journal of Educational Psychology. 26: 654, 1935. 22 Robert G. Bei*nreuter, "Validity of the Personality Inventory," Personnel Journal. 11: 383, 1933. 54 correlations from a number or studies average about ten points lower, or .75. These coefficients Indicate high Internal consistency for the test, although they are not high enough for accurate individual p r e d i c t i o n . 23 Patterson has also summarized, in this same study, numerous other studies concerned with the relation of the Bernreuter scores to the following variables: (1 ) ratings and case studies, (2 ) various clinical groups, (3) intelligence, scholastic aptitude and achievement, (5) social groupings, (4) special abilities, (5) family resemblance and birth order, (7) physical and physiological factors, and (8 ) other personality test scores. Patterson reports correlations, ..hich he considers to be low, thought ranging as high as .78* In each instance he reports negative or inconclusive findings. Summary. Regarding the research done with the Bernreuter P e r s o n a l i t y Inventory, the following statements appear to summarize the findings as pertinent to this study: Cl) Within limits, the Personality Inventory can be used to learn something of the adjustment of college students. (2) The P e r s o n a l i t y Inventory is a usable research instrument in some situations. There appear to be situations 23 Cecil llm Patterson, "The Relation of Bernreuter Scores to Parent Behavior, Child Behavior, Urban Rural Residence, end Other Background Factors in One Hundred Normal Adult Parents," Journal Q? Social Psychology. 24: 5, 1946. 55 in which the Parsnnp.litv Inventory is not usable principally because or poor conditions of test administration* (3) The Personality Inventory (though not as reliable as intelligence tests) yields reliability coefficients comparable with other tests of personality, being best estimated by a reliability coefficient of .85. Correlations with other similar personality tests can probably be termed "fairly high*" IV. THE CEAEHORE JvULALURES OF MUSICAL TALENT Seashore jfeflgUfag. s£1 MuaJLcal T.algfifc. This test is one of the oldest of the musical talent tests and is probably the best known. forms: It consists of e battery of six tests in two (1) Series A, fox* use with undifferentiated groups and, (£) Series E, for use with musical groups. are given by means of the phonogx*aph. These tests The original battery Included the following tests: (a) Sense of Pitch: This test consists of one hundred pitch comparisons of varying difficulty, the subject being required to decide whether a second tone is higher or lowex* than the first. (b) I n t e n s i t y p; fsqpimination: This consists of one hundred comparisons of two tones differing more or less in intensity, the subject being required to decide whether the second is louder or softer than the first. (c) Sense of Time: This consists of one hundred comparisons of the length of two time Intervals marked off by clicks. (d) Sense of C o n s o n a n c e : This consists of fifty comparisons between pairs of two-tone clangs, the subject being required to judge whether the second clang Is better or worse than the first on the basis of smoothness. 56 purity, and blending. (e) Tnn«i Memory: This consists or riTty comparisons between two sets or unrelated tones# one tone In the set being changed on repetition, the subject being required to identiTy the changed tone. (r) Sense or Rhythm: This consists or Tifty comparisons between pairs or rhythm patterns, the subject being required to Judge whether the second pattern is the same as, or dlTferent rrom the rirst.24 In 1959, Seashore and others published a manual oT instructions to go with their revision or this battery. Lewis says regarding the revision or the Seashore Measures: OT the six tests in the original Seashore battery (pitch, intensity, time, rhythm, tonal memory, and consonance), all have been retained except the consonance test. Each one has been modified tp a considerable extent but is basically unchanged.25 The rollowlng summary or research done with the Seashore • Measures of Musical Talent, consists or studies done with the original battery. In a study investigating the validity or the Seashore Measures or Musical Talent. Mursell outlines the rollowlng points: (1) Reliabilities were round by correlating the rirst and second rive rows or the scoring tables and applying the Spearman-Brown prophecy rormula. . . . Reliabilities so obtained were approximately similar to those or previous studies. . . . Reliabilities as re­ vealed by the standard error were so low that the use or ^ Don Lewis, "The Timbre Test in the Revised Seashore Measures," Journal oT Applied Psychology. 25: 108, 1941. 85 Ibid.. 25: 108. 67 the tests Tor Individual diagnosis seemed questionable* (2 ) In most or the tests, the performance of conservatory students was superior to that of college students, measured both by averages or median overlapping* This superiority was not sufficiently marked to warrant any very specific educational advice* (3) Grades in Applied Music, and estimates of musical talent seemed to have a reliability which should render them good validation material* There was almost no relationship between Seashore Test performance, and the above criteria (applied music grades and talent ratings )• Taken with similar results from other studies, this leads to the conclusion that the tests cannot make fine discriminations of true musical talent within musical groups* The relationship of the Seashore Test scores to performance on special tests indicates that the former may be of use as aids in diagnosing special musical abilities* Where significant correlations between Seashore Tost scores and tests of special abilities and types of achievement were found, we are usually dealing with very heterogeneous groups, which indicates that the battery may be able to discriminate roughly, though it cannot do so very accurately or finely*26 Mursell reports the following reliability coefficients: Pitch Intensity (Loudnes s ) Time Consonance Tonal Memory Rhythm r •66 .86 N 161 164 .81 .52 .88 •64 164 165 165 163 These coefficients are comparable to those reported by Seashore and somewhat higher than other studies* 27 Mursell also reports correlation coefficients between the Seashore tests and talent ratings, piano final grades and James L* Mursell, "Measuring Musical Ability and Achievements: A Study of the Correlations of Seashore Test Scores and Other Variables," Journal Eflnewtinnni fififfflftCfifr, 25: 125, 1932. 87 Ibid., 25: 118. 58 voice final grades* All of these coefficients are low r&iiging from a negative correlation of *27 to a positive correlation of .25. hone of these correlations are as large as three times its probable error* Somewhat higher correlations have been found between Seashore test scores and other special ability tests* Gaw* 28 for 149 cases, foimd the following correlations for her sightsinging test with the beashore Tests: .36; Tonal Memory, *56* Pitch, *46; Intensity* Kosher29 obtained the following coefficients between his group measures of sight-singing and the Seashore Tests: Time, *3581, Consonance, *2912, Pitch, *4391, Tonal Memory, *4386, and Intensity* *485* These corx-elations were found for numbers varying from 430 30 to 460. Wright* for 24 cases, found correlations bat ween her three sets of music achievement tests and the Seashore battery as a whole of *45, *51, and *73* In a similar study of the Seashore tests, McCarthy makes the following general statement: "The Seashore tests Esther Allen Gaw* "Five Studies in the Music Tests," Psychological Monographs, 39: 145-156, 1928. 29 n . K. i'osher, £ Study oC the Or pud Method s L Measurement of Sleht-Sln,:lt«.. Contrlbutions to Education, rio. 194* KewYork: Teachers College, Columbia University, 1925, 75 pp. 30 F. A. Wright, "The Correlation Between Achievement and Capacity in Music," Journal Educational Research* 17: 50-56, 1928. 59 are of the greatest practical value at the extremes or the scale and as Is the case with most measures of vocational aptitude, It Is much easier to predict failure than It Is to pz*edlct success*“31 Additional studies concerned with the original Seashore Measures of Musical Talent will be found in Chapter Two, “Summary of Pertinent Literature," of this study* Ih& 1232. Revision Talents * Sfifl&ilQEfe Measures o£ Jfejglfifll The records on which these measures are recorded are described by Saetvelt, Lewis, and Seashore: These records are twelve-inch double-faced Victor records. They were recorded in the RCA recording laboratory and are available In all music houses and RCA Victor dealers handling Victor records In this country and abroad. They consist of two series* Series £ Is designed for use with unselected groups, as in the schoolroom, or in general group surveys. This series furnishes a general dragnet for the discovery and rating of six different talents. The £L Series Is designed for testing of musical groups or individuals, as in the selection for musical organi­ zations , admission to music schools, the assignment to musical Instruments, or In the search for causes of failure in music* Series JB may also be used for individual measurement where greater reliability Is desired, as in the music studio or the psychological laboratory* Each of the two series, A and B, measures the same talents and each series consists of three doublefaced records which may be purchased separately; but for general use both series should be available. Series A covers the full range of talent or lack of ^ Dorothea McCarthy, "A Study of the Seashore Measures of Musical Talent," Journal of Applied Psychology. 14: 454, 1930. 60 talent and is therefore essential for such surveys. Series B covers a narrower range and is therefore more diagnostic and more economical in musical situations. There ere three records for each series, with one measure on each side. The records are listed as follows: Series Record No. 450A 450 B 451 A 451 B 458 A 458 B A Test Pitch Loudness* Time Timbre Rhythm Tonal Memory Series B Test Record No. 453 A Pitch 453 B Loudness* 454 A Time 454 B Timbre 455 A Rhythm 455 B Tonal Memory *The Acoustical Society of America has defined and recommended the use of the term "loudness"to designate what we have formerly called ^intensity. Seashore writes the following description of this 1939 revision: These measures present the following characteristics: they are based on a scientific analysis of musical appreciation and performance; they deal with elements which function in all music; they are standardized for content so that alternate or new series are not needed; they give quantitative results which may be verified to a high degree of certainty; they are economical in that expensive instruments are replaced by phonograjtti records; they may be used with any language and at any racial or cultural level; they are simple and as nearly selfoperative as possible; they are designed for group measurements; they are Interpreted in terms of established norms. They ore called measures to distinguish them from the ordinary paper and pencil tests and because they are patterned on principles of accurate measurement with 30 Carl £• Seashore, Don Lewis, and Joseph G. Saetvelt, OL Instructions and Interpretations for £&£. frft&SAgre IS. of Musical Talent. 1939 Revision. Camden, New Jersey: iucational Department, RCA Victor Division, Radio Corporation of America, 1940, pp. 4-5. 61 scientific instruments in the laboratory. They are based upon two fundamental laws of scientific measurement. The first Is that the factor under observation must be Isolated in order that we may know exactly what we are measuring. This factor is varied under control while all other factors are kept constant. Thus in measuring the sense of time* we vary duration only, keep all other factors constant, and avoid complex situations. The second principle maintains that the conclusion to be drawn must be limited specifically to the implications of the factor which has been measured under control. Thus if we measure the sense of rhythm and find a very superior performance, the conclusion is not that the subject is musical; it is merely that the individual has a very superior sense of rhythm. . . . they (these measures) may be used extensively as class experiments in general psychology, music, and phonetics. They are convenient measuring tools for acoustical research in many fields. They do not measure training or achievement in music. Excellence in these is a condition for artistic appreciation and skills in performance; but it does not in Itself guarantee such achievements. They do not measure intelligence, feeling, or the will to work. They do not furnish a single, all-inclusive index to musical ability. They should not be averaged; each score is but an item in the musical profile. They are not fool-proof. As measuring instruments they are fully adequate, but the use of them requires tact, skill, ability to motivate* favorable atmosphere, and wisdom in interpretation.53 The general procedure used in revising the Seashore Measures of Mu r Iq a I T a Iant is described by Saetvelt and others as follows: The first step in the process of revising was to make an item analysis of the original measures on the basis of the responses made to each item (or group of items) by large numbers of school children and adults. The analytical data yielded information on the relative 33 Ibid.. pp. 3-4 68 difficulty of each item end served as a basis for choosing items to be included in tentative or trial forms. These trial forms were recorded on acetate records and admin­ istered to three different age groups: fifth- and sixthgrade pupils, seventh- and eighth-grade pupils, and adults. The results were used for a second evaluation of individual items. A new set of tentative forms was then recorded and administered. This general procedure was followed until items covering a satisfactory range of difficulty had been selected for both Series A and Series B of each of the new measures.34 The following coefficients of reliability for the revised measures together with the means and standard deviations of the scores obtained on each measure by large groups of subjects are tabled by Saetvelt. 35 Pitch A Loudness A Time A Timbre a Rhythm A Tonal Memory A Total No. 1071 1037 1116 852 1104 980 Mean Score In Percentiles 75.9 31.4 76.3 75.1 83.5 83.3 Pitch E Loudness 3 Time B Timbre B Rhythm B Tonal Memory B 752 777 792 603 794 731 69. G 76.9 66.3 60.9 71.8 70.4 S. D. 12.2 10.8 9.7 9.7 9.2 13.5 r .88 .88 .76 .74 .62 .83 P. E .01 .01 .02 .02 .02 .01 10.6 11.1 10.1 8.2 11.3 17.5 .78 .77 .70 .72 .72 •89 .02 .02 .02 •02 .02 •01 These measures are published with a manual of instructions and interpretations by Seashore and others in which norms are available. These norms make it possible to convert the number 34 Joseph G. Saetvelt, Don Lewis, and Carl E. Seashore, Revision the Seashore Measures o£. TflJLglfta* Iowa City, Iowa: university of Iowa Press, 1940, p. 13. 35 Ibid.. p. 34. 63 of right answers into a ranking order- "The highest rank, 1, represents the scores of the highest ten per cent in a normal unselected community- The next highest, 2 , includes scores made by the next ten per cent and so on to rank ten which includes scores made by the lowest ten per cent of e normal population-" V. MUSICAL ABILITY RATINGS The ratings of musical abilities- Ratings of the following musical aspects were included in this study: (1 ) kusicality, (2) Ability to Sight-Read music, and (3) Ability to Perform- In order to obtain ratings of these abilities for these 130 students, the help of the Music Faculty of Michigan State College was enlisted. The process was implemented in the following manner. First, the general outline of the study was i^resented, verbally to this faculty group- The general rating procedure was outlined at this meeting. Secondly, a letter was sent to each staff member out­ lining the purpose and technique of the study. As regards to the ratings, the purposes of this letter were: (1) To acquaint each music staff member with the musical abilities for which a rating was needed- ^ Seashore, op. clt-, p- 16- This letter Included 64 a variety of definitions for each musical ability. Each faculty member was asked to state his reaction to these definitions as to adequacy and pertinency. In this way* consensus definitions were arrived at for each musical ability. (2) To give each faculty member opportunity to list those individual students for whom ratings based upon adequate contact could be made. An alphabetical checklist of students was included with this letter. Applied music teachers were askeu to rate their private students in applied music as to iiusicality, Eight-Heading, and Ability to Perform. Teachers of class instruments and voice classes were asked to rate their students on these three abilities as they applied to the instrumental or vocal music class. Theory teachers were asked for ratings of J&asicality for their theory class students. Other faculty members were asked to check those students for whom ratings could be made, based on adequate contacts with the student in situations other than those listed above. As these checklists of students were returned, a second letter was sent to each staff member with ( l ) a discussion of the rating procedure and (2 ) individual rating scale forms for each student included on the particular checklist. This second letter included the following; of the rationale of the scales. (1) Statement "Each of the rating scales has many factors which contribute to it in some unknown and not directly measurable ratio. It would seem most satisfactory 66 and Justifiable, therefore, to obtain the desired rankings through the Judgment of professionally qualified persons* It is assumed, further, that each item will be considered as a single concept rather than as a composite of discrete elements* ** (2 ) Detailed statement of the procedure to be used in making these ratings. The rating instructions were made explicit according to procedures delineated by Adkins* 3 7 (3 ) The final form of the definitions of abilities in the rating scales as follows: x Having a responsiveness to music; having a fondness or intelligent appreciation for music; saving a sensitivity to musical feeling; having an inner urge towards music* Ability to Sight-read* Ability to perform music of a reasonable grade of difficulty at sight on his major or minor instrument. Ability to perform music of a reasonable grade of difficulty at sight on his class instrument or in the vocal class. In general: Ability to organize musical material into an intelligible performance at sight* Ability to Perform. Ability to organize studied or memorized music into a musical performance commensurate with 37 Dorothy Adkins, Construction aadL Anftlvg&g J2L AfiM.gy.gr, ment Tests. JVashington, D. C. : United States Government Printing Office, 1947, p. 179. 66 his level of training and musicality before any audience* In general; Ability to realize his musical potentialities in a performing capacity before any audience. (4) Rating scale forms for individual students. These forms included information as to the student's major instrument, curriculum, year in school and age at which he began music study. (See the Appendix for the actual rating form and the discussion of the rating procedure.) VI. APPLIED 2.1USIC GRADES AP.BJ.AftA m&ls. ££&£& averages. Each music student Is required to enroll for a minimum of two years study of applied music at Michigan State College. In most instances undergraduate students continue applied music study with the same instrument throughout all four years of college. This applied music study consists of private lessons on a musical instrument 02* in voice. It appeared safe to assume that grades received for such work would be of some usefulness to the present research problem. Letter grades are used at Michigan State College and are given the following quality point values: A B C D F - 4 quality points 3 quality points 2 quality points 1 quality point no quality points 67 In order to arrive at a workable score relating to the applied music study grades the following procedure was used: (1) Grades received during consecutive school terms for the 1950-1951 regular session were selected for inclusion in the study* This Included grades received In Fall quarter 1950, .Tinter quarter 1951, and Spring quarter 1951. cases, consecutive grades were not available* In three Consecutive grades for the previous session were used In these Instances* In one Instance, grades were not available, because the level of work done was below college requirements* (2) A score combining grades received In each of these three consecutive terms was arrived at by converting letter grades to quality points and summing. This procedure yielded "averages" which were readily usable and which avoided the use of decimals* CHAPTER V SUMMARY OF STUDY DATA This chapter presents the test data in various forms pertinent to the problem. The chapter presents data designed (1 ) to point up the performance of this group on the several tests used in the study, (2 ) to show the relation between this test performance of the group and ratings and grades, and (3) to point up the test performance of several sub-groups. For the sake of conciseness, certain abbreviations have been used throughout this chapter and those to follow. I. ABBREVIATION’S FOR TESTS The American Council on Education Psychological will be abbreviated as follows: ACE will refer :o the test in general, ACE-T to the total score, ACE-Q to he Quantitative score, ACE-L to the Linguistic score. The Eernreuter is follows: Inventory will be abbreviated BPI will refer to the test in general, Bl-N to he Neuroticlsm score, B2-S to the Self-Sufficlency score, >3-1 to the Introversion score, B4-D to the Dominance score, l-C to the Confidence score, and F2-S to the Sociability core. 69 The C&QjjegfttiLye Reading Comprehension Test will be abbreviated as follows: CRCT will refer to the test in general, CRCT-T to the Total score, CRCT-V to the Vocabulary score, CRCT-R to the Speed of Heading Score, and CRCT-C to the Level of Comprehension Score. The scores of the Seashore jfeasaggg. of. M as leal T fluent will be Identified simply as the Seashore scores. The various measures will be Identified as Pitch, Time, Timbre, Rhythm, Loudness, and Tonal Memory tests. The musical ability ratings and applied music grades will be Identified as such. II. THE ACE RESULTS Graph 5 presents the decile distribution of ACE-T scores for the total group of students studied. These scores tended toward negative skewness, with a concentration of scores above the fifth decile. This distribution tended toward bl-modallty with modes at the tenth and fifth deciles with 28 and 23 scores, respectively. Above the fifth decile, there were 61.04 per cent of the scores, whereas 50 per cent would be expected. Above the eighth decile, there were 30.23 per cent of the scores, whereas 2 G per cent would be expected. The middle four deciles had 43.60 per cent of the scores and the bottom three deciles contained 15.11 per cent of the scores. The decile mean of these ACE-T scores was 6.37 with a standard 71 deviation of 2.67. The decile having the most scores (the mode) was the tenth* with 23 scores, the total. 16*27 per cent of It should be noted that this decile had more scores than the sum of those scores In the first, second, and third deciles. Graph 6 shows the distribution of the ACE-L scores. These scores exhibited a tendency, similar to that of the ACE-T scores, toward negative skewness. Above the fifth decile, there were 63.37 per cent of the scores, with 17.44 per cent of the scores clustered at the tenth decile. The lowest three deciles taken together, contained 16.27 per cent of the scores, fewer than In the tenth decile alone. The middle four deciles contained 44.12 per cent of the scores. The decile mean for this distribution of ACE-L scores was 6.40 with a standard deviation of 2.97. This distribution exhibited considerable tendency toward eveness of distribution of scores above the fourth decile. Graph 7 shows the decile distribution of the ACS-Q scores. This distribution showed somewhat less negative skewness than was found In the other ACE distributions. The middle four deciles taken together contained 40.11 per cent of the scores; the bottom four deciles contained only 27.32 per cent of the scores. The mean of the ACE-3 scores was 6.08 with a standard deviation of 2.61. Of these scores, 59.81 per cent were above the fifth decile, with 50.58 per cent above the sixth decile. wr~ •* ■ - Dpcllej aoores Graph rajp 6 DJSTRlBUTilON Op AOB-L SCORBS 30 25 20 15 10 Decile scores Graph 7 DISTRIBUTION OP ACB-iQ SCORBS Table I gives the decile means of the ACE scores made by the four undergraduate classes* Seniors as a group scored slightly higher on all ACE scores than the other three classes* All averages on the table were higher than the firth decile. In Table II, the decile means scored by the several curricular groups are shown* The Theory, Science and Arts, and Music Therapy students had the highest averages on all ACE scores. The larger groups tended to show small differences between groups. In general, the ACE-L averages were higher than the ACE-^ averages. All averages were above the fifth decile. On Table III, the ACE means are shown computed for the several major Instrument groupings. The averages for the woodwind instrument players were highest In each Instance, with the averages of the string instrument players second highest. In general, the ACE-L averages were higher than the ACE--i averages. This was not true, however, when there were more males than females in a grouping, as in the Instance of brass Instrument players, who had higher ACE-Q averages. All averages were above the fifth decile. On Table IV, averages scored on the ACE are given for the male and female students. The ACE-T averages showed only a slight difference In favor of the males. The males had the highest *.CE-Q average, while the females had the hignest ACE-L average. fifth decile. All averages were higher than the 75 TABLE I A V E R A G E ACE SCOR E S FOR THE FOUR CLASSES Class______ No._______ A C E - L ______________ A C E - Q _______________ ACE-T ^'reshraan 66 6*42 5*91 6 .29 .opLotnor© 28 6.59 5.43 6.14 Junior 45 6.22 6 .09 6.24 Senior 34 6.71 6.79 7 .00 76 T A B L E II A V E R A G E ACE SC OKIE.' FOR T H E CUREICUI.AR GROUPS *1 1 R! O < ACE-2 ACE-T 6.32 8.25 5.89 71 6*30 5.94 6.28 39 6*05 6.07 6.15 16 7.80 7.18 7.56 11 7.72 6.64 7.55 5 7.60 7.00 7.60 Curriculum No • Applied fcuslc 28 general* Ins t r u m e n t a l : Theory clence & Arts iuaic T h e r a p y *SK.«— School r u s i c 77 T A B L E III A V E R A G E ACS SCORED FOR TFS INSTRUMENTAL GROUPS Instrumental Gr oup No. ACE-L A C E — Q. a c p :-t Brass 18 5.33 5.61 5.44 Woodwind 28 7.04 7.00 7.18 string 14 b* 71 6.71 6.93 iriano 69 6.77 6.25 6.64 Voice 42 6.24 5.26 5.93 T A B L E IV AVERAGE ACE SCORED FOR TJ'E MALE AND FEMALE GROUPS Sox No, ACE-L AC S— ACE-T sale 66 6*24 6* 50 6.44 107 6*53 5.78 6.36 Female 79 Table V shows product-moment correlation coefficients found between the ACE scores and ratings of lluslcallty, Sight-Reading ability* Performance ability, and applied music grades. These ACE scores were not significantly correlated with Performance ability ratings nor with applied music grades. The coefficients of correlation found between ratings of Ihisicality and Sight-Reading ability were found to be signi­ ficant, though low. The correlations involving the ACE-Q scores and ratings were larger than those correlations involving the ACE-L scores and the ratings. single correlation was .305 The largest 07 found between ACE-T and Uusicality ratings. III. THE CRCT RESULTS Graph 8 presents the distribution of CRCT-T scores by deciles. This distribution was similar to those of the ACE scores, exhibiting a tendency toward negative skewness. The distribution was a multi-modal one with concentrations of scores at the thii*d, fifth, seventh, eighth, and tenth deciles. The tenth decile had 29 scores, or 16.86 per cent of the total, and in the top five deciles were 61.62 per cent of the scores, whereas, 50 per cent would normally be expected. The middle four deciles contained 37.79 per cent and the bottom three deciles contained 21.51 per cent of the total. The mean of this CRCT-T distribution was 6.05 with a standard deviation of 2.65. 80 TABLE V C O K H E L kTION C O E F F I C I E N T S P O UND B ETWEEN ACE SCORES AND T H E M U S I C A L A B I L I T Y RAT I N G S A N D A P P L I E D KUSIC GRA D E S i-.O - Scores Kualcality Reading Performance Mualc Gradea ACE-L .218±;07 • 213 ±.08 •01 2 £ . 0 8 .082i*.07 ACE- L .271L.07 •250±.07 .095t.08 •002±.07 ACE-T •305^.07 • 242±.07 •0 3 9 ± . 0 8 .074 ±.07 Number of | oases 35 30 15 lO Dec11a scores _ __ : _ _ Graph. & DISTRIBUTION 0& CRCTi-T SCORES 10 88 The distribution of the CRCT-V scores is shown on Graph 9. This distribution of scores was bi-modal, but exhibited considerable negative skewness* In the top five deciles there were 63*37 per cent of the scores, with 38*55 per cent concentrated in the ninth and tenth deciles, taken together* In the middle four deciles were 41.27 per cent of the scores and IS.oe per cent of the scores were found in the first three deciles taken together. The mean of these scores was 6.55 with a standard deviation of 2*41* The distribution of CRCT-R scores is shown on Graph 10. The distribution was negatively skewed and had modes at the third, the sixth and seventh, and tenth deciles. In the upper five deciles, there were 61*04 per cent of the scox*es with 29.06 per cent in the top two deciles, taken together* The middle four deciles contained 37*79 per cent and the bottom two deciles, taken together, had 12.20 per cent of the total. The mean of this CRCT-R distribution was 6.31 with a standard deviation of 2 .86* The distribution of the CRCT-C scores is given in Graph 11. This distribution tended toward negative skewness but was unique in its tendency toward flatness* Mo well marked inodes appear, although there was a piling up of scores above the sixth decile. In the top four deciles there were 51.74 per cent of the scores and 38.37 per cent were found in the middle four deciles, and 22.67 per cent were found 5 4 6 e 7 0 Decile score> Qrapli 9 DISTHIBUTIQH JQF GRCT-V SCORES t Bumbe# of o«.s*s t tT- 4- - i f k 4 30 - 25 20 IS 10 ♦ 5 i 3 4 5 6 Decile score* 7 Graph lO DISTRIBUTION OF CRCT-R SCORES 8 9 10 r- : 4 5 6 Dejelle soorao [orapb lY [ 7 DXSTRIBDT^OH 09 GRC’D-O SQORES *1 4 _. -t - h I ... i •i U-' )*'•i l *• rl- . U - . . - i - t -f- i 86 in the bottom three deciles. The mean or the distribution was 6.19 with a standard deviation or 2*89. The means of each or the CRCT scores was higher than the fiTth decile. In Table VI, the means scored by the rour undergraduate classes ror each or the CRCT scores are shown. were ranked by CRCT-T averages as rollows: sophomores, juniors, and freshmen. The classes seniors, There was no definite profile pattern from class to class on these averages. All averages were above the fifth decile. In Table VII, the distribution of CRCT means (in deciles) is shown computed for curricular groupings. The averages of the Music Therapy, Theory, and Science & Arts groups tended to cluster from one to two deciles above the Applied, PSM-» general, and PSM-instrumental curricula. fitted all curricula groups. Ho profile pattern All averages were above the fifth decile. In Table VIII, the CRCT averages are given for major instrument groupings. Average differences between the groups were not large except in the instance of the brass Instrument group, which had averages approximately a decile or more below the other groupings. In general, the averages of the woodwind instrument group were highest on these CRCT scores. All averages were higher than the sixth decile, with the exception of those of the brass instrument group. 87 TAILS VI AVERAGE CRCT SCORES FOR THE FOUR CLASSES kpeed or Comprehension Level of Comprehension "lass No« Vocabulary Total Fresh. 65 6.40 5.99 6.12 6.14 Soph. 28 6.54 6.68 5.79 6.39 Jr. 45 6.20 6.23 6.44 6.31 'r • 23 6.76 6.27 6.21 6.64 68 table vii AVERAGE CRCT SCORES FOR THE CURRICULAR GROUPS Curriculum Speed ol1 Level of No. Vocabulary Comprehension Comprehension Total App. Music 28 6.64 6.39 6.28 6.42 SK-general* 71 6.21 6.08 6.18 6.24 39 5.97 5.53 5.20 5.66 Theory 16 7.56 7.10 7.43 7.50 science & Arts 11 6.91 7.00 6.27 7.00 5 7.20 7.40 7.40 7.80 instru­ mental Music Therapy *SM— School Music 89 TABLE VIII AVERAGE CRCT SCORES FOR THE IN.STRULENTAL GROUPS T&pwrar of "LeVeT'o'f-----instrument No* Vocabulary Comprehension Comprehension Total * • CD ' 4.39 4.94 6.79 6.61 6.68 6.43 6.21 6.36 6.50 69 6.59 6.45 6.44 6.59 42 6.59 6.07 6.19 6.34 Brass 18 8.56 Woodwind 88 6.36 string 14 Piano Voice 4 90 In Table IX, the CRCT means are shown computed for sex groups# The female students had higher averages in all instances, with the exception of the CRCT-V scores# betYjeen the groups were not great# Differences All averages were above the fifth decile. The correlation coefficients found between the CRCT scores and the musical ability ratings and applied music grades are shown in Table X. coefficients on the table: There were two significant (1) between the CRCT-V scores and the Musicality rating, #267 :fc#07; and (2) between the ChCT-R scores and the Jftisicallty ratings, .249^.07. IV. THE BPI RESULTS The BPI scores were computed in percentiles and were used In this form throughout the study except wheii compiled into frequency distributions. For clarity and conciseness, decile scores were used in the distributions of Bl-N, B2-S, 33-1, and B4-D scores# In Graph 12, the distribution of the Bl-N scores is shown by deciles# This distribution was multl-raodal, but had considerable positive skewness# There were 69.07 per cent of the scores in the bottom five deciles, with 47.61 per cent In the bottom three deciles and 35#91 per cent In the bottom two deciles. The middle four deciles contained 35.91 per cent of the scores# The percentile mean of these scores was 38.91 with a standard deviation of 26.56. 91 TABLE IX AVERAGE CRCT SCORES FOR TEE MALE AND FEMALE GROUPS ' Vocabulary Spaed or Comprehension Level of1 Comprehension ex No> Total Male 66 6.57 6.09 5.78 6.20 Female 107 6.42 6.50 6.36 6.42 92 TABLE X CORRSL h TIOSi COEFFICIENTS POUND BETWEEN CRCT SCORES AND THE MUSICAL ABILITY RATINGS AND APPLIED MUSIC GRADES ST'sHB-' Readinft ” CRCT Scores Muslcallty ferformanoe -AppTleH" ~ Music Grades CRCT—V .2671.07 .206±.08 .102 ±.09 .1421.08 CRCT—R .2491.07 .222±.08 •094±.08 .1171.08 CRCT—C .1331.07 .1491.08 •057±.09 .0151.08 CRCT—T .2011.07 •2151.08 .lOOt.08 .1111.08 so 20 IS XO Decile score* jOrapli 12 j j DISTHjXBUTION OP BERHREUTER PER SO NAlilTY INVENTORY Bl-N SCORES i 10 94 The distribution of the B2-S scores is shown in Graph 13. This distribution resembled a normal distribution in several ways. (1 ) The distribution had a marked central tendency, with 33 scores at the sixth decile. (2 ) The top five deciles contained fifty per cent of the scores and the middle four deciles contained 52.22 per cent of the scores. The percentile mean of these scores was LI.52 with a standard deviation of r-> # O Off (• The distribution, in deciles, is given for the B3-I scores in Graph 14. This distribution showed the sharpest negative skewness of all the BPI distributions, but showed some resemblance to the Bl-N distribution. In the bottom five deciles were 76.66 per cent and in the top two deciles were 42.22 per cent of the scores. In the top two deciles were only 4.44 per cent of the scores. The percentile mean of these scores was 31.79 with a standard deviation of 25.28. The distribution of B4-D scores is shown on Graph 15. The distribution was bi—modal, but was similar in several respects to a normal distribution. The top five deciles contained 48.33 per cent of the scores. The middle four deciles contained 52.22 per cent of the scores and the middle two deciles contained 31.66 per cent of the scores. In the bottom four deciles were 35.55 per cent of the scores. The percentile mean of the distribution was 48.29 with a standard deviation of 24.60. 1 2 I a 4 § 6 7 8j Decile scfereis •Orapfia I •- ; : *■ DIBT eU b HTION OF RSBKREUTKR PERSONALITY tMVBIf&ORX 02-S SCORES . "4" . . T B 6 oil*I adore BITCIQIL EHVBHV 3CORB3 LJ^S Qzfaph 16 QP BKRNRTfTiy 5R-LE INVBNPORY BOHA 98 The distribution of Fl-C scores Is shown on Graph 16. These scores were all concentrated within four deciles ranging from the 53rd percentile to the ninety-third percentile* These scores were graphed In percentiles (grouped In Intervals of two) to show the sharpness of the bl-modallty of the distribution* From the fifty-second percentile to the seventy-seventh percentile were 52.22 per cent of the scores in a distribution having a strong central tendency* The area between the seventy-eighth percentile and the ninetyfifth percentile contained 46*11 per cent of the scores. This mode had a concentration of 63 scores between the eightysecond and ninety-first percentiles showing a second strong central tendency, In this total distribution of Fl-C scores* The median score, of the mode found between the fifty-second and seventy-ninth percentiles (the lower mode), was approxi­ mately 69. The median score, found between the seventy- eighth and ninety-fifth percentiles, (the upper mode ), was approximately 87* These modes were separated by 18 percentiles* The mean of the total distribution was 75*32 percentiles with a standard deviation of 10*59* The median score of the total distribution was at the seventy-sixth percentile. The distribution of the F2-S scores is given In Graph 17. These scores were graphed in percentiles with a single percentile in each interval* All of the scores were above the eighty-fifth percentile with 90.55 per cent above the ; ! I r I 1 05 r-i t Q i T C ^ O > r - 4 * C « Jsl Ol ioi. in -.ch oa h io if> t> r-~ o c^ ch ad a a) a) go ift -tfj-io 1C tO-’O t£)i toi Per^Gftti.-'6'"WC. q> 1 j { 1 njbet-of c^asffl 89 90 91 92 95 Percentile scores; j j Graph 17 .teiisotwi: IHVEKTQRYi F2-S SCORES - -i. ; 1 — H 8 101 ninetieth percentile, and 51.11 per cent above the ninety-third percentile. The distribution showed a marked modal tendency at the ninety-fourth percentile which had 58 scores, 32.22 per cent of the total. Despite a strong modal tendency, the distribution simulated a bell-shaped distribution. The mean or these scores was 93.08 with a standard deviation of 1.95. Table XI gives the averages of the BPI scores computed b classes. Averages for all classes on all scores tended to cluster around the total group averages. described the profile of these averages. A single pattern The averages of these BPI scales tended to cluster as follows: Bl-N averages around the thirty-ninth percentile, the B2-S averages around the fifty-second percentile, the B3-I averages around the thirty-second percentile, the B4-D averages around the fortyeighth percentile, the Fl-C averages around seventy-fifth percentile, and the F2-S averages around the ninety-third percentile. The averages for the BPI scales are shown on Table XII computed for the several curricular groups. The profile pattern described by each of these curricular averages was described as follows: Bl-N, low; B2-C high; B3-I, low; B4-D, high; Fl-C, very high; and F2-S, extremely high. Sharpest differences from scale to scale were found In the profile of averages of the Ifuslc Therapy group, consisting of five students. The BPI averages of the larger curricular lo e TABLE XI AVERAGE BrI SCORES FOR THE FOUR CLASSES Class_____ No, Bl-N B2-S B3-I 134-I) Fl-C F2-S 68 39*16 50.07 31.44 47.14 74.86 92.97 ;oplioinore 29 40.97 56.03 33.59 50.03 75.86 93.41 Junior 47 38.81 53.28 31.96 49.81 73.62 93.23 Senior 36 37.53 50.94 31.06 49.06 77.97 92.88 'reshman 109 TABLE X I I A V H/vGIS B r l SC 3HK: . FOR TV.Z CUR IIC U L A R GROUPS Curriculum No* Bl-N 32- S 33-1 34—D Pl-C R2-S Applied ruelc 50 58*43 57.63 30. 66 51.55 72.70 95.70 :)lf*general<- 72 55*65 49.57 28.26 48.90 70.85 92.08 .‘V—lna truraental 42 44*90 42.57 37.35 42.66 84.57 92.21 Theory 16 47*24 65.56 41.62 50.68 82.06 94.18 'cienoe & Arts 12 39.91 55.17 25.92 52.67 70.92 93.67 5 10.00 73.60 7.40 77.80 64.40 93.40 iusle Therapy oM— School )-uelc 104 groups had smaller differences, (1 ) between groups and (2 ) between scales, than those of the smaller groups* Averages for the BPI scales are shown on Table XIII commuted for the several major instrument groupings. The sharpest profile pattern was found in the averages for the string instrument students. The string and piano students had scale averages separated by at least a decile from those of the other groups. In general, profile patterns were similar from group to group. The BPI averages for the male and female students are shown on Table XIV. The profile pattern from test to test was sharper for the females than for the males. These profile patterns were similar, however. Table XV gives correlation coefficients found between BPI scores, musical ability ratings, and applied music grades. The coefficients were all very low, none significant. The ratings and grades appeared to be unrelated to the BPI scores. V. THE SEASHORE TEST RESULTS The scores for the Seashore tests are computed as inverted decile scores.1 For clarity and consistency, the frequency distribution of the Seashore scores have been 1 Carl E. Seashore, J. G. Saetvelt, and Don Lewis, Manual .&£ Instructions AQd BeSs^.. T.rB«y« “ uoatloMl Department, RCA Victor Division, Radio Corporation of America, 1939, 19 pp. 105 TABLE XIII AVERAGE BPI SCORES FOR THE INSTRUMENTAL GROUPS Instrument No. Bl-N B2-3 B3-I B4-D Fl-C F2-S Fraa s 18 47.05 47.06 42.33 45.28 86.17 92.56 Woodvrind 30 43.80 46.37 36.47 44.00 81.23 92.83 string 15 32.40 56.00 29.13 54.33 76.73 92.67 Plano 72 37.53 56.14 30.49 51.21 72.11 93.50 Voice 43 38.64 49.21 31.48 46.81 72.33 93.00 106 T a BLK XIV AVERAGE BE I SCORES FOR THE i'AIJi AND FNi-iADE GKOUES Sex No* Bl-N B2-S B3-I B4-D Fl-C F2-S 2 . ale 71 46.44 47*30 37*28 46.86 86.11 92.44 109 34*86 56*14 27.46 51.35 68.29 93.52 Female 107 TABLE XV CORRELATION COEFFICIENTS FOUND BETWEEN BPI SCORES AND THE MUSICAL ABILITY RATINGS AND APPLIED MUSIC GRADES Applied Music Grades BPI Soores Musicality SightReading Bl-N •005±.05 .044±.08 *097±.08 — *028±.07 B2-r •107±.05 -.030±.08 .0284*.08 •149±.07 B3-I .0021*.06 .022 lr.08 •10C±»08 «.02S±.07 B4-D •130 ±.07 •054+.08 .043+.08 •023±.07 Performance 108 graphed In the normal order with the top decile as ten and the bottom decile as one* In Graph 18 is shown the distribution of the Seashore Pitch scores by deciles* was very sharply skewed negatively* This distribution In the top five deciles were Q8 per cent of the scores, with 57.98 per cent in the top two deciles* In the bottom five deciles were 27*30 per cent of the scores* The mean of these scores was 7*57 with a standard deviation of 2*70. The distribution of the Seashore Loudness (Intensity) is shown on Graph 19* This distribution exhibited negative skewness combined with a tendency toward flatness above the fourth decile* In the top six deciles were 85.80 per cent of the scores, with 70.41 per cent in the top five deciles and 36*09 per cent In the top two deciles. In the middle four deciles were 32*54 per cent of the scores and in the bottom four were 14.20 per cent of the scores. The mean of these scores was 7*11 with a standard deviation of 2*49* The distribution of the Seashore Time scores is shown in Graph 20. The distribution was negatively skewed with 78.10 per cent of the scores in the upper five deciles and 74*55 per cent in the upper four deciles* deciles were 26.83 per cent of the scores* In the bottom six The mean of these scores was 7.39 with a standard deviation of 2.52* The distribution of the Seashore Timbre scores is shown on Graph 21* These scores were the most sharply skewed of 25 eo 15 lO lO i ~ - f— j : D^elle scores ; ; 1 rarapS^lB"] I i r r~ • DISTRIBUTION p? SE^SgOR^ PITCH a # KbxamaiHisxir- ct ot 03 ■93""t o$- r ov !©«(©*© 30 ltob11* ftc*i*oi8 D^clSle Isaore Grj&pn 21 oi SfaA^HOftg B SC 113 all the several scores In the data with 90.53 per cent of the scores in the upper rive deciles and 55.OB per cent In the tenth decile alone. In the bottom five deciles were 10.05 per cent or the scores. The raean of these scores was 3.55 with a standard deviation of 2.17. The median score was approximately at the ninety-first percentile. The distribution of the Seashore Rhythm scores Is shown on Graph 22. The distribution was sharply skewed negatively, with 87.57 per cent of the scores In the top five deciles and 64.49 per cent In the top two deciles. In the bottom five deciles were 23.07 per cent of the scores. The mean of these scores was 3.20 with a standard deviation of 2.19. The median score was approximately at the eighty-fourth percentile. The distribution of the Seashore Tonal Memory scores is shown on Graph 23. Sharp negative skewness was exhibited, with 87.55 per cent of the scores In the upper five deciles and 47.92 per cent in the tenth decile alone. four deciles were only 4.73 per cent. In the bottom The mean of these scores was 8.34 with a standard deviation of 2 .01. The median score was found at approximately the eighty-ninth percentile. The scores of the six Seashore tests discussed above were averaged and referred to as a total, or "T", score. (The meaning of this average score Is not entirely clear. u» - 6 I 7 Decile sejere s Grfepjr 2(3 F|IBtJT|0N iOff SEASjHORE jTORAh WEMOpY SCORES ’T I „ . .1 t- . • a.. -J. f■ .i T* ■f- i r4: •t ■ —i-i -t- J : •--*r- t'1 r . [.I-: ,U- mn t; ■ 1X6 Seashore g . points out that these scores are not to be averaged* Several studies have used such averages, however, and for purposes of comparison, this average was computed and used in the study. ) The distribution of these Seashore averages is shown on Graph 84* The distribution was a negatively skewed one* avex-age scores were below 3.6 deciles* No The scores were heavily concentrated above the sixth decile and tended toward a sloping skewness from the eighth decile to the third decile. The mean of these scores was 7.05 with a standard deviation of 1.81. In Table XVI the Seashore averages scored by the four class groups are compared* Inspection of the table revealed no strong ranking pattern or large differences in class averages* From class to class and from Measure to Measure scores were fairly high and uniform. Seniors had the highest average on three tests; Juniors, Sophomores, and Freshmen had the highest average on a single test each* The Seniors had the highest six-test average. Table XVII shows a comparison of the averages scored by the various curricular groups* The SM - Instrumental students tended to score highest, with highest averages on three single tests and the highest six-test average score. 2 Ibid., p. 4. t ■ - :—I — •-t—r ■H- ; - . ! ■ 'T-' :~4~~ » T » . i : j t i -i i . : . i t *. i . i 1 - - I ....... J $ 'Dwells aooreja : » f Graph 24 • • - . i rHE■■ ■BXS-TB3T f c PJOJt OF 3 KVHRAOBSGOHIS __ -.mi2. ■ - 'f: •t . . 4:: . i■ j ’ 44 , - - 118 TABLE XVI AVERAGE SEASHORE SCORES FOR THE FOUR CLASSES Class No. Pitch Loud­ ness Rhythm Time Tonal Timbre Kemory 7.17 6.73 3.71 7.31 8.39 Soph. 28 7.86 6.25 8.25 7.93 Jr. 42 7.96 7.55 7.89 r. 34 7.89 7.92 7.48 8.47 7.80 8.43 * O • CO Fresh* 62 Averaj 7.81 7.29 8.08 8.15 8.01 7.24 8.95 8.68 8.05 119 TABLE XVII AVERAGE SEASHORE SCORES FOR THE INSTRUMENTAL OROUPS LoudTonal neaa Rlyythm Time Tlm~bre Memory Average 17 8.70 7.64 8.59 8.00 8.53 8 • CO Woodwind 29 8.28 7.21 7.97 8.21 8.48 8.52 8.11 String IS 8.13 6.47 7.93 8.13 8.87 7.80 7.92 Piano 70 7.34 6.99 8.25 CO • <0 8.29 8.45 7.67 Voice 39 6.69 7.31 8.03 8.33 8.13 7.63 • H 00 Brass CD • C* * Instrument No. Pitch 120 The Theory students were highest on two single tests, but also had the lowest single test average. Scores tended to be fairly high and uniform, however, from curriculum to curriculum. The distribution of Seashore averages computed for the various instrument groups is shown in Table XVIII. The averages for the voice students tended to be lowest as indicated by a six-test average of 7.63 deciles. The brass instrument students tended to score highest, with highest averages on four single tests and with the highest six-test average score. The woodwind instrument stridents scored highest on two single tests and had the second highest sixtest averages. This comparison shows that, in general, the brass, woodwind, and string students, in that order, tended to score higher on these tests than the piano and voice students. The Seashore test averages computed for male and female students are shown on Table XIX. Differences in averages for these groups were not large. Each group had higher averages on three tests; the six-test average favors the male group slightly. The correlation coefficients found between the Seashore scores and the musical ability ratings and applied music grades are given in Table XX. All correlation coefficients with the applied music grades and performance ratings were low and 121 TABLE XVIII AVERAGE SEASHORE SCORES FOR THE CURRICULAR GROUPS Curriculum No* Fitch TowS-----Tonal ness Rhythm Time Timbre Memory Average App. Music 30 7*77 7.53 7.37 7.50 8.43 8.33 7.85 SM-general 68 7.02 7.14 8.46 7.11 8.43 8.19 7.74 SM-instru­ ment a1'* 40 8.12 7.45 8.02 8.37 8.62 8.12 8.12 Theory 16 8.18 5.50 8.88 7.13 8.06 8.75 7.77 Science & Arts 10 7.00 6.40 8.60 6.20 8.10 8.20 7.44 5 7.60 6.40 8.80 7.60 8.60 9.00 7.98 Music Therapy *SM— School Music lbe TABLE XIX AVERAGE SEASHORE SCORE ■ FOR THE EALE AND FE; .ALE GROUPS Sex ale Female Pltoh Eou5^ Tonal nesa Rhythm Time Timbre fcemory Average 69 7.74 6.98 7.87 7.57 8.52 8.25 7.91 101 7.35 7.02 8.33 7.33 8.32 8.35 7.81 Ko« 123 T a BLE XX CORRELATION COEFFICIENTS FOUND SETiYEEN SEASHORE SCORES AND THE MUSICAL ABILITY RATINGS AND APPLIED teUSIC GRADES 'oaabor« keaaures U & i - Laaicallty Heading Performance Applie< Fusic Grades Pitch .2161.07 .2111.08 •215 ±.08 .143 ±.07 Loudness .215 ir.07 •1871.08 .0121.08 .1161.07 -.0341.08 •003 S’.08 -.1011.08 -.1171.08 .0691.08 .1511.08 •059±.08 •0171.08 -.0141.08 .065±.08 •176±.08 .0501.08 Tonal Nersorjr .1891.07 .210±.08 .0611.08 •0591.08 Average Score .2301.07 .254±.07 .1031.08 .0741*07 Rhytha TLre Tlnbre 124 not significant. (That is, the coefficient was smaller than three times its standard error. ) The coefficient found between the six-test averages and the Sight-Reading rating was significant, .25 4:.07, but low. A significant correlation was found between the six-test average and the Muslcallty rating, .230 £.07. Between the Pitch and Loudness scores and the Muslcallty ratings, coefficients of .216.1:.07, and *213±. .07, respectively, were found. This table was characterlzed by low coefficients, consisting of four significant co­ efficients and twenty coefficients which were not significant. Three of the four significant coefficients were between feashore scores and Muslcallty ratings. VI. R3SULTG OF ThE MUSICAL ABILITY RATINGS AND APPLIED MUSIC GRADES The distribution of the ratings of Musicality, (grouped by scale units) are shown on Graph 25. The distribution resembled a normal distribution but was slightly negatively skewed. Tnere were 30.49 per cent of the ratings in the top three scale intervals (7, 8, and 9), designated "upper third"). In the middle three scale intervals (4, 5, and 6, designated "middle third" ) were 59.15 per cent of the ratings. In the bottom three scale intervals (1, 2 , and 3, designated "lower third" ), were 10.36 per cent of the ratings. The mean rating (in scale units) was 5.98 and the standard deviation of the distribution was 1.51. 185 Grtaph. 85 F AVERAGE RATINGS OF IfOSIOAjLITY 126 The distribution of* the Sight-Heading ability ratings is shown on Graph 26. The distribution resembled a normal distribution but tended toward flatness near Its mean* ratings were distributed as follows: The In the upper three scale units were 24*14 per cent, in the middle three scale units were 64*14 per cent, and in thebottom three scale unitswere 11*72 per cent of the ratings* The mean of these ratings in scale units was 5*68 with a standard deviation of 1*51* The distribution of the Performance ability ratings is shown on Graph 27* This distribution was roughly bell-shaped, but had a slight negative skewness* The ratings were distributed as follows: In the upper three scale units were 29*93 per cent of the ratings, in the middle three scale units were 62*04 per cent, and in the bottom three sclIg units were 8*03 per cent of the ratings. The mean of these ratings, In scale units, was 5*96 with a standard deviation of 1*45* Table XXI shows the distribution of the averages of the musical ability ratings computed for the four classes* The classes were ranked on each scale as follows: Seniors, Sophomores, Juniors, and Freshmen* Comparison of average musical ability ratings for the various curricula is shown on Table XXII* The applied music students had the highest average rating on each scale* The Theox*y students had second highest averages on each scale; HUBlbO D 3 M RIBOTIOJ# OP SI ABILI 40 35 30 25 ~T .« «?, .•« : O; O Rating"b b 41s |unttp Graph 217 tj p' | ;' ■ OR ins rRiBDiEioH- pp \ay OP PJRPORMAHCK rrrt 1S9 TA BLE XXI AV-'Ra GE R .TINGS .)F m u s i c a l a b i l i t y FOR THE FOUR CLASSES Class No* Muslcallty Sight-Reading Performance Freshman 59 56.00 57 .16 57.00 Sophomore 27 62. 81 60.04 61.08 Junior 44 59.27 54.21 58.79 ..enlor 35 63.66 60.90 62.43 130 TABLE XXII AVERAGE RATINGS OP MUSICAL ABILITY FOR THE CURRICULAR GROUPS Curriculum No* A p p l i e d Music 30 stl-generol- Sight-Reading Pe r f o rmance 6 5 .78 64.04 70.48 72 57*38 55*29 56.14 SK-1 na t rums nt al 42 59*59 57*98 59.41 Theory 16 64*92 60.27 65.64 Science & Arts 12 4 9 *63 43.00 44.33 5 54*80 5 2 .60 53.60 M u sic T h o r a p y "*SM——School Music I.iuslc all ty ■the Sk - Instrumental students had third highest averages on each scale; the SM - general students had fourth highest averages on each scale; the Huslc Therapy students had fifth highest averages in each instance and the Science and Arts students had the lowest averages in each instance. The ranking of the curricular averages was the same on each rating scale. A comparison of the averages of the musical ability ratings made by students of the several instrument groupings is shown on Table XXIII. Differences did not follow a set pattern, but in general, the brass, woodwind and string Instrument students had averages higher on each scale than those of the piano and voice students. The string students scored highest on the performance ability scale. The musical ability rating averages for the male and female students are shown in Table XXIV. The male students scored higher averages on all three scales; differences, however, were small. The distribution of the applied music grades is shown on Graph 28. These grades were distributed as follows: grades of "A", 17.85 per cent; grades of "B", 44.64 per cent; grades of "C", 32.73 per cent; and grades of "D", 4.76 per cent. This distribution tended to be normal despite the misleading shape of Graph 88. The averages used in the Graph were quality point sums and exhibited of course, less central tendency than would be expected of a grade distribution* 132 TABLE XXIII AVERAGE RATINGS OP MUSICAL ABILITY FOR THE INSTRUMENTAL GROUPS Instrument No. Brass 17 Woodwind. Slglit-Roadlng Performance 64.71 60.89 61.17 30 59.47 59.52 60.24 String 13 60.77 58.54 67.44 Plano 65 57.75 54.13 58.82 Vole© 30 59.69 58.91 57.44 Muslcallty 133 TABLE XXIV AVLRAiliS R A T I N G S OP L U S I C A L A B I L I T Y FOR T H E R A L E A N D FLf-iALE GROUPS lex_____ No>______Muslcallty____ Sight-Heading Performance Male 67 61.55 58.61 62.35 Female 98 58.29 56.57 57.45 S3 OS I • • sejsttO JO 135 The applied music grade averages are shown on Table XXV as scored by classes* The classes were ranked by this average in the following order: (1 ) Seniors, (2 ) Sophomores, (3) Juniors, and (4) Freshmen. Table XXV APPLIED MUSIC GRADE POINT AVERAGES FOR FOUR CLASSES SI fig3_____ Freshmen Sophomores Juniors Seniors Grade Average Q. 719 9.393 9.068 9.771 28 42 34 Applied music grade averages are shown in Table XXVI by the several curricular groups. These curricular groups were ranked from highest to lowest averages as follows: Applied music, SM - Instrumental, Music Therapy, SM - general Science and Arts, and Theory. Table XXVI DISTRIBUTION O F APPLIED MUSIC GRADES BY CURRICULAR GROUPINGS Group Applied SM - Instrumental SM - general Theory Science &. Art Therapy No. 30 42 68 10.83 8.83 8.69 14 8.21 10 8.30 8.80 5 These grade averages are shown in Table XXVII as scored by the major Instrument groupings. These groups were ranked from highest to lowest averages as follows: (1) Voice 136 students, (2) String students, (3) Brass instrument students (4) Woodwind instrument students, and (5) Piano students. Table XXVII DISTRIBUTION OF APPLIED MUSIC GRADE POINT AVERAGES FOR TilE INSTRUMENTAL GROUPS Instrument Piano Woouwlnd Brass String Voice 70 32 19 16 41 8.48 8.87 9.31 9. 60 10.20 The female students had an applied music grade average of 9.43; the male students had an average of 8.82. VII. The ACE results. SUMMARY OF RESULTS All distributions of ACE scores for the total group were negatively skewed, with means above the sixth decile. The profile pattern tended toward a high ACE-L score and a slightly lower ACE-Q score. From freshman to senior class there were slight differences in ACE averages. Seniors had the highest averages; the other three class groups had generally lower averages but Indicating no clear-cut ranking pattern. The smaller curricular groups (Theory, Music Therapy, and Science <1 Arts) had higher ACE averages than the larger curricular groups. The averages of these smaller groups were clearly higher (a decile, or more) than th»CE averages* Sex differences on the ACE scores were not large and were most clearly apparent In profile patterns. The profile pattern of the male students had an ACE-Q average higher than the ACE-L average. The reverse of this pattern characterized the averages of the female students. Correlation coefficients showed some relation between the ACE scores and the ratings of Muslcallty and of SightReading ability. Apparently, the ACE scores were not related to the ratings of Performance ability nor to applied music grades • The CRCT r e s u l t s . The distribution of the CRCT scores were all negatively skewed, in varying degrees. Each total group average of these scores was higher than the sixth decile. The various classes were ranked according to the magni­ tude of the CRCT-T average as follows: Seniors, Sophomores, Juniors, and Fi'eshmen. decile or higher. All averages were near the sixth No consistent differences were shown. The Theory, Science and Arts, and Music Therapy students tended to have the highest averages of the curricular groupings on all of the CRCT tests. The SM - Instrumental students had the lowest averages of the curricular groups. 133 The CRCT averages or the major Instrument groupings showed only small group differences except In the Instance of the averages of the brass Instrument group. These latter averages tended to cluster around the fifth decile or below. All other averages were above the sixth decile. Sex differences In CRCT scores were small, but generally favored the female students. The CR C T scores showed some relation to the musical ability ratings, but no relation to the applied music grades. The CRCT-C scores were less closely related to these ratings than were the other CKCT scores. The BPI r e s u l t s . This group of students tended to score slightly low on the Bl-N scale (thirty-ninth percentile), higher on the BS-S scale (fifty-second percentile X low on the B3-I scale (twenty-fifth percentile), higher on the B4-D scale (forty-eighth percentile), considerably higher on the Fl-C scale (seventy-fifth percentile), and extremely high on the F8-S scale (ninety-third percentile). '.Vhen BPI averages were computed for classes, this same profile pattern (referred to above) was shown. i/hen BPI averages were computed for curricular groupings, these groups were ranked, according to the range of between test differences, as follows: Music Therapy, Theory, Applied Music, Science and Arts, SM - general, and SM - Instrumental. 139 The profile patterns of the 3k averages tended toward flatness the SM - Instrumental group averages tended to be the reverse of that for the whole group# The profile pattern of the BPI averages, computed for the major instrument groupings, was similar to that of the group as a whole* The profiles of BPI averages of the brass and woodwind Instrument students tended toward flatness* The sharpest profiles of averages were those of the string and piano students. The musical ability ratings and applied music grades appeared to be unrelated to the BPI scores* The ^.Afihofe r e s u l t s . The distributions of the Seashore scores were all negatively skewed, some sharply so* Means were all above the seventh decile; three were above the eighth decile* When median scores were computed, these scores were usually higher than the mean. When averages for the Seashore scores were computed for classes, these averages showed small differences, without a well-marked trend* Seniors had highest averages on three tests and on the six-test average* Juniors, Sophomores, and Freshmen followed in that order* The curricular groups were ranked by Seashore averages as follows: SM - instrumental, Music Therapy, Applied Music, Theory, SM - general, and Science and Arts* Differences were not great and showed no clear differentiation pattern* 140 When averages for the Seashore scores were computed for the several major Instrument groupings, the brass Instrument group tended to have averages slightly higher than those of other groups* Averages for this group were higher on Tour or the tests and on the six-test average* Next, in order or ranking, were the woodwind, string, piano, and voice groups* Differences between the Seashore averages of the male and female students were small and not consistent* The correlation coefficients between the Seashore scores and the musical ability ratings and applied music grades were low* Low, but significant coefficients were found between (1 ) Pitch scores and Kuslcallty ratings, (2 ) Loudness scores and Uuslcallty ratings and (5) Six-test average scores and Ifusicality and Sight-Reading ability ratings* The rest of the Seashore scores appeared to be only slightly related to, or not related, to these ratings and grades, as indicated by coefficients which were not significant* X&s. mfiXsaJ, flteUJLto rgfiulta* The distribution of the musical ability ratings tended to resemble normal distributions* The distributions of these ratings were, however, slightly skewed, negatively* On both the Uuslcallty and Performance ratings, the classes were ranked by class averages as follows: Sophomores, Juniors, and Freshmen* Seniors, On the Sight-Reading 141 scale the rankings were Senior, Sophomore, Freshmen and Juniors. On all ratiiig scales, curricular groups were ranked by averages, from highest to lowest, as follows: Applied, Theory, Eli general, S14 instrumental, Music Therapy, and Science and Arts. Differences between adjacent rankings were not large. Differences between the highest and lowest averages in each scale, varied from one to more than two standard deviations. There were small differences in the rating averages of the various instrument groups with no clear ranking pattern □etwaen groups nor between ratings indicated. The averages for the male students were higher* than hiose of the females on all of the rating scales. T.fts Akhl. led m s X c fifftde results. By applied music ^raae averages, the classes were ranked as follows: Sophomores, Juniors, and Freshmen. :he curricular groups as follows: Seniors, These averages ranked Applied, SM - instrumental, ftislc Therapy, Sl£ - general, Science and Arts and Theory. ?he several instrumental groups were ranked by grade averages is follows: voice, string, brass, woodwind and piano. The ipplied grade average of the female students was higher than ;hat of the male students. The applied music grades appeared :ot to be related to any of the scores of the four test lotteries used. CHAPTER VI SIGNIFICANCE OF THE STUDY DATA In Chapter V, the data of the study were factually presented. Chapter VI points up the significance of these iata including the following points: Cl) the significance af the performance of the group studied on the four tests itilized in the study; (2 ) the relation of the test scores ;o the musical ability ratings and (3) the Implications of ;he test profile of the musical group studied. I. SIGNIFICANCE OF THE ACE RESULTS The performance of the group under study on this test ras clearly a superior one. Each of the ACE means for the otal group was found to be statistically significantly different from a mean of 5.5 deciles (taken as the best iStimate of the average of the population that has taken he test )• The ACE-T mean. Statistical tests of the significance f the difference between the ACE-T mean of 6.37 for this roup and estimates of the mean of the large group of students hat have ever taken this ACE test were made. The difference etween ACE-T mean of 6.37 deciles (for the group studied) 143 and an estimated population mean of 5*50 deciles was found to ae significant beyond the one per cent confidence level. When S.03 was taken as the population mean, this difference was 3till significant at the one per cent confidence level. The ACE-L mean. Tills mean (6.46 deciles) was found to ie slgnifIcantly different from an estimated population mean >f 5.50 deciles, beyond the one per cent confidence level. ?he critical region of this confidence level was not reached mtil the estimated population mean was raised to 6.1b deciles. Xhs. Afifi-a m s m - This mean (6.08 deciles) was found to >e slgnifIcantly different from an estimated population mean »f 5.50 deciles, beyond the one per cent confidence level. 'he critical region of this confidence level was reached when he estimated population mean was raised to 6.01 deciles. II. THE SIGNIFICANCE OF THE CRCT RESULTS All of the CRCT means for the whole group were found to e slgnifIcantly different from a mean of 5.5 deciles beyond he one per cent level of confidence. The CRCT-V mean. This CRCT-V mean (6.55 deciles) for he group was found to be slgnificantly different from an stlmated population mean of 5.5 deciles, beyond the one per ent confidence level. The critical region of this confidence 144 .evel was not readied until the estimated population mean ras raised to 5*97 deciles. The CRCT-R mean. This mean (6.25 deciles) was round o be slgnificantly different from an estimated population jean of 5.50 deciles beyond the one per cent confidence level. *he critical region of this level of confidence was not reached mtil the estimated population mean was raised to 6.06 deciles. The CRCT-C mean- This mean (6.11 deciles ) was found to >e slgnificantly different from an estimated population mean >f 5.50 deciles beyond the one per cent confidence level. 'he critical region of this confidence level was not reached mtil the estimated population mean v;as raised to 6.07. The CRCT-T mean. This mean (6.05 deciles) was found to >e significantly differentfrom an estimated population mean >f 5.50 beyond the one per centconfidence level. The critical ‘egion of this confidence level was not reached until the estimated population mean was raised to 6.02 deciles. III. THE SIGNIFICANCE OF THE BPI RESULTS The differences between estimated population means and observed means of the BPI scores were not significant In tvery Instance. The means of the B2-S and B4—D scores were ound to be not signif icantly different from estimates of 145 *,he population mean. The means of the other scales (Bl-N, 33-1, Fl-C, and F2-S) were found to be significantly different, iome highly so, from estimates of the population mean beyond ;he one per cent level of confidence* The Bl-N mean. The Bl-N mean (26.56 percentiles ) was 'ound to be very highly significantly different from an estimated population mean of fifty percentiles beyond the >ne per cent confidence level. The critical region of this jonfidence level was not reached until the estimated population lean was lowered to 31.56 percentiles, a probably untenable estimate of the population mean. The B2-S mean. The mean of the B2-S scores (51. 52 >ercentiles ) was not found to be signif icantly different ‘rora an estimated population mean of fifty percentiles. The tritical region of the five per cent level of confidence was tot x'eached until the estimated population mean was lowered o 46.42 percentiles. The B3-I mean. The B3-I mean was found to be slgnif icantly ifferent from an estimated population mean of 50.00 deciles ieyond the one per cent level of confidence. The critical eglon of the confidence level was not reached until the stlmated population mean was lowered to 36.64 percentiles, . probably untenable estimate of the population mean. 146 The B4—D mean. This mean (48*29) was round to be not significantly different from an estimated population mean of fifty percentiles at the one per cent level of confidence. The critics- region of this confidence level was reached when 53.01 percentiles was taken as the estimate of the population mean. The Fl-C mean. This mean (75.32) was found to differ slgnificantly from an estimated population mean of fifty percentiles beyond the one per cent level of confidence. The critical region of this level of confidence was i*eached when the estimated population mean was raised to 72.95 percentiles, a probably untenable estimate of the population mean. The F2-S mean. This mean (93.08) was found to differ slgnificantly from an estimated population mean of fifty percentiles beyond the one per cent confidence level. The critical region of this confidence level was reached when the estimated population mean was raised to 89.01 percentiles, a probably untenable estimate of this mean. IV. THE SIGNIFICANCE OF THE EEAEHOHE RESULTS All of the Seashore means for the whole group were found to differ slgnif icantly from an estimated mean of 5.00 deciles Three of these means were at least two deciles, and three, were three deciles above this estimate, a difference which was found to be highly significant well beyond the one per cent level of confidence. 147 V. RELATION OF MUSIC i*BILITY RATINGS TO TEST SCORES It has been shown that the test performance of this group of music students in several instances was slgnificantly different from the expected norms of test performance. The following discussion Is concerned with product-moment correlation coefficients found between these several test scores and the ratings of musical abilities. An examination of the correlation tables In Chapter V revealed the following facts. Cl) These several test scores appeared to be unrelated to applied music grades, and to ratings of performance, (2 ) The ACE scores appeared to be slightly related to ratings of musicality and of EightReading ability. Of the six correlation coefficients found between the ACE scores and the ratings of Musicality and of Sight-Reading ability, five were significant correlations, beyond the one per cent level of confidence. The average of the coefficients found between the ACE scores and the Musicality ratings was .265. The average of coefficients found between these scores and the Sight-Reading ratings was .198. (3) The CRCT scores appeared to be slightly related to the ratings of Musicality and of Sight-Reading ability. Of the eight co­ efficients found between CRCT scores and the ratings of 146 Musicality and or Sight-Reading ability; two were significant and four coefficients approached statistical significance at the one per cent confidence level. The average of the co­ efficients found between the CRCT scores and the Musicality ratings was .213. The average of the coefficients found between the CRCT scores and the Cight-Reading ratings was •198. (4) The coefficients found between the above tests (ACE and CRCT) tended to be homogeneous, and varied but slightly from coefficient to coefficient. (5) The coefficients found between the Seashore scores and the Musicality and Eight-Reading ratings tended to be heterogeneous, varying both in size and sign. Of the fourteen coefficients found between the Seashore scores and the Musicality and SightReading ratings, four were significant at the one per cent level of confidence and two coefficients approached statistical significance at this level. (6 ) All of the significant coefficients tended to be homogeneous with a difference between the largest and smallest coefficient of only .09, and with a range from .211 between Seashore Pitch scores and Sight-Reading ratings, to .306 between ACE-T scores and Musicality ratings. The coefficients found between the ACE and CRCT scores and musical ability ratings were more homogeneous and more consistent than were coefficients found between Seashore scores and these ratings. 149 Zhfe %SS.%. profile g£_ the grous. The profile of test scores for the group of znuslc students studied* revealed the following facts. (1 ) The average scores of this group on the ACE were all above the sixth decile* with the ACE-L scores tending to be slightly higher than the AGE-Q scores. This difference between the ACE-L scores and the ACE-Q scores tended to be reversed for the male music students. The observed difference between these two ACE scores for the total group was apparently due to the fact that there were more female students than male students. (2) Average scores of this group on the CRCT were all above the sixth decile. A majority of the scores were above the estimated test population mean on all sub-scores. (3) Average scores on the Seashore single tests and on the six-test average were all above the seventh decile; three were above the eighth decile. On all sub-tests* these scores tended to concentrate at the ninth and tenth deciles. (4) Four of the BPI sub-score averages were significantly different from estimated averages of unselected college groups. On the scales of Neurotlclsm* Introversion* Confidence* and Sociability* these differences were significant beyond the one per cent level of confidence. (L) Curricular groups were found to have related rankings on the ACE and CRCT tests* as Indicated by rank difference correlation coefficient of .63 found between 150 rankings or the ACE-T and CRCT-T averages* The rankings or the CRCT-T averages were related to the Bl-N and B3-I means els Indicated by rank difference correlation coefficients of-.94 and -*83 respectively. (6 ) Instrumental groups had related rankings on the rtCE-T and CRCT-T means as indicated by a rank difference correlation coefficient of *90* Other related rankings were Lndlcated by the following rank difference correlation co­ efficients: (a) between Seashore six-test average means and <’1-C means, .90; (b) between Seashore six-test average means and F2-S means, .85; and (c) between jUUslcallty ratings means m d applied music grade average means, .80* The following :*ank difference correlation coefficients were found between Sight-Reading rating means and BPI means as follows: Ca) 31-N means, -.90; (b) B2-E means, .90; (c) B3-I means, -.90; [d) B4-D means, .80; and (e) Fl-C means, .90* soefflcients found were: Other large (a) between Bl-N means and B3-I leans, 1.00; (b ) between Bl-N means and B4-D means, -.90; c) between B2-S means and B3-I means, -.80; (d) between 32-S means and B4-D means, .90; (e) between B2-S means and '1-C means, .80; (f) between B3-I means and B4-0 means, -.90; ind Cg) between Fl-C means and F2-3 means, -.90. (7) Certain rankings of class group means appeared to >e related. The relation of these rankings between ACE-T ind BPI means was indicated by the following rank difference 151 correlation coefficients: between AGE-T and: (a) Bl-N, .80; (b) B8-S, .80; Cc) B3-I, 1.00; 11*383-386, 1933. _____ . "Chance and Personality Inventory Scores," .Tmimfti aL Educational Psychology. 26:279-283, 1935. Blenstock, Sylvia P., "A Review of Recent Studies on Musical Aptitude," J&WTiaJL £ L Educational Psychology. 33:427-442, 1942* Broom, M. E., "A Note Concerning the Seashore Measures of Musical Talent," School end Society. 30*274-275, 1929. Burton, Martha V., "The Effect of College Attendance Upon Personality as Measured by the Bemreuter Personality Inventory," Journal of Educational Research. 38*708711, 1944. Christensen, Arnold M., "Traits of College Going, Employed, and Unemployed, and High School Graduates," School aa.YAfiH> 46:597-601, 1938. Coleman, James C. and Jean Elizabeth McCanly, "Nail-Biting Among College Students," JgarapA °£ Abstract and Social Psychology. 43:517-525, 1948. Cureton, E., "Note on the Validity of the American Council on Education Psychological Examination," Journal of AfiPXAS& P-fiZgh.Ql.Qfiy.> 23*306-307, 1939. Dodge, Arthur E., "Social Dominance of Clerical Workers and Sales-Persons as Measured by the Bernreuter Personality Inventory," Journal q £_ Educational Psychology. 28:71-73, 1937. Embree, R. £•, "Note on the Estimation of College Aptitude Test Scores from Intelligence Quotients Derived from Group- Intelligence Tests," Journal SlL Educational Psychology. 37*502-504, 1946. Feder, D. D., and L. Opal Baer, "A Comparison of Test Records and Clinical Evaluations of Personality Adjustment," ifeMXAflJL ££ 32*33-44, 1941. Fisher, Willis, and Samuel P. Hayes, Jr., "Maladjustment in College Predicted by Bernreuter Inventory Scores and Family Position," Journal of Applied Psychology. 25*8696, 1941. 16X Flanagan* John C., "Technical Aspects or Multi-Trait Tests: A Reply to Dr. Lorge," Journal, ££. gffyCkftAOEy.. 26:641-651, 1935. Gaw, Esther Allen, "Five Studies in the Music Tests,*' Pgy.S.hQlQfil.gfti MOflPfifftBhg* 39:145-156, 1928. Hauser, Luellen J. Munn, "A Comparative Study of the Intelligence of University Freshmen Enrolled in Business and Liberal Art Schools," Journal OL 43:49-57, 1949. Hildreth, Gertrude, "Comparison of Early Binet Records with College Aptitude Test Scores," Journal of Educational Psychology. 30:365-371, 1939. Hollingworth, L. S., "lAisical Sensitivity of Children Who Test Above 135 I. Q . J o u r n a l of Educational Psychology. 17:95-109, 1926. Hunter, E. C., "Changes in Scores of College Students on the American Council on Education Psychological Examination at Yearly Intervals During the College Course," Jyiypa,\ J2£ Re.se flESil, 36:284-239, 1942. Jarvie, L. I., "Does the Bernreuter Personality Inventory Contribute to Counseling," Educational Research ffuAAeftla, 17:7-9, 1938. Lanlgan, Mary A., "The Effectiveness of the Otis, the ACE and the Minnesota Speed of Reading Tests for Predicting Success in College," Journal o£ SdUOfttAPflftl flggQftrCll. 41:290-296, 1947. LaycocM, Samuel R., "The Bernreuter Personality in the Selection of Teachers," Educational Artminiafcr*t:4 n n SUPgffYAPlftH, 20:59-63, 1934. Lewis, Don, "The Timbre Test in the Revised Seashore Measures," Journal of Applied .Psychology* 25:108-112, 1941 Livesay, T. M., "Does Test Intelligence Increase at the College Level, " Journal qL Educational Egy.gji9.l9ig.> 30:63-68, 1939. Lorge, Irving, "Personality Traits by Flat: A Correction," Journal of Educational Psychology. 26:652-654, 1935. . "Personality Traits by Fiat: I. The Analysis of the Total Trait Scores and Keys of the Bernreuter Personality Inventory, Journal p£. nn*1 Fay 26:273-278, 1935. 168 ■ "Personality Traits by Flat: II* The Consistency of the Bernreuter Personality Inventory by the Bernreuter and Flanagan Keys." Journal ££. fiHieatinnm PgygfrglflfiY., 261487-434, 1935* McCarthy, Dorothea, "A Study of the Seashore Measures or Musical Talent," Journal of Applied Psychology* 14:437455, 1930. McPhall, Andrew N«, "Quantitative and Linguistic Scores on the American Council on Education Psychological Exam­ ination," School and Society.* 56:248-251, 1942. Martin, Glenn C., "A Factor Analysis or the Bernreuter Personality Inventory," E d u c a t i on a l and psychological Measures. 8:85-92, 1948. Miller, Richard, "Uber liuslkallsche Begabung und lhre Besiehungen za Sonstigen Anlagen," Zeitschrift fur £Sy-fiUgjLfi&te. 97:191-214, 1925. Munroe, Ruth L., "Rorschach Findings on College Students Showing Different Constellations of Subscores on the American Council on Education Psychological Examination," Journal fi£. Consulting Psychology. 10:301-316, 1946. Mursell, James L., "Measuring Musical Ability and Achievement: A Study of the Correlations of Seashore Test Scores and Other Variables," Jnur»nal g£. ftgP.gaf.Sh> 25:116-126, 1932. . "Intelligence and Musicality," Education. 59:559562, 1939. Nemxek, Claude, "The Value of the Bernreuter Personality Inventory for Direct and Differential Prediction of Academic Success as Measured by Teachers' Marks," Journal o£ Applied .P.sy.chflAfifi2L> 22:576-586, 1938. Pannenborg, H. J., and W. A., "Die Psychologle des Muslkers," £M£. FSYfthgAflflAft. 73:91-136, 1915. Patterson, Cecil H., "The Relation of Bernreuter Scores to Parent Behavior, Child Behavior, Urban Rural Residence, and Other Background Factors In One Hundred Normal Adult Parents," Journal of Social Psychology. 24:3-49, 1946. Peterson, B. H., "Can You Predict Who Will Be Able to Do College Work," School Executive. 61:26-27, 1941. 163 Reed. Homer B., "The Place or the Bernreuter Personality Inventory, Stenqulst Mechanical Aptitudes and Thurstone Vocational Interest Tests In College Entrance Tests," Journal of Applied Psychology, 25:528-534, 1941. Ryans, David 0., "A Tentative Statement or the Reliabilities oT Persistence Test Scores to Certain Personality Traits as Measured by the Bernreuter Personality Inventory," SqgittftTY £&& JsaauO, &L genetic Psychology. 54:229-234, 1939. St. Clair, Walter F., and J. Conrad Seegers, "Certain Aspects oT the F Scores or the Bernreuter Personality Inventory," Journal nf Ediifffitf 9171*3, Psychology. 29:301-311, 1938. Schmitz, Sylvester B ., "Predicting Success in College," Jour&al S L Educational Psychology. 28:465-473, 1937. Seashore, Carl E., "Musical Intelligence," Music Educators Journal. 24:32-33, 1938. ________ , "Talent," School and Society. 55:169-173, 1942. Seder, Margaret, "The Reliability and Validity or the American Council on Education Psychological Examination, 1938 Edition," Journal of Educational Research, 34:90-101, 1940. Smith, D. D., and Frances 0. Trlggs, "Educational Success and Failure of Students with High Quantitative and Low Linguistic Scores on the American Council on Education Psychological Examination," American Psychology. 5:353354, 1950. Stalnaker, Elizabeth M., "A Four Year Study of the Freshman Class or 1935 at the "Vest Virginia University," J o a m \ or Educational Research. 39:81-101, 1945. Stogdill, Emily L., and Minnie E. Thomas, "The Bernreuter Personality Inventory as a Measure of Student Adjust­ ment," Journal aT Social Psychology. 9:299-315, 1938. Super, Donald E., "The American Council on Education Psychological Examination and Special Abilities," Jon^nfii or Psychology. 9:221-226, 1940. "The Bernreuter Personality Inventory: A Review of Research," teXShPlQ&teoJL 39:94-125, 1942. 164 Votaw, David F., ”Regression Lines Tor Estimating Intelligence Quotients and American Council on Education Examination Scores,’* Journal o£ Educational Psychology. 37:179-181, 1946. Wallace, W. L., ”DIfferential Px*edictlve Value of the American Council on Education Psychological Examination,** School and Society. 70:23-24, 1948. Ward, Lewis B., and Samuel A. Kirk, **studies in the Selection of Students for a Teachers College, ” Journal, of Educational Rc86arch. 35*665—672, 1942. Wheeler, Lester R., ‘•Summary of a Study of the Intelligence of University of Miami Freshmen, ** J o u r n a l of Educational Research. 43:307-308, 1949. Willett, G. MReliability of Test Scores on Seven Tests,” Journal of E d u c a t i o n ^ Rescax'ch. 43:293-298, 1949. Wilson College Studies in Psychology, ”A Comparison of the Wechler-Bellevue, Revised Stanford-Blnet, and i-raerican Council on Education Tests at the College Level,” Journal Of Psychology. 14:317-326, 1942. Wright, Frances A., ”Tlie Cox*relation Between Achievement and Capacity in Music,” Journal fill Educational j&§ejg££h, 17:50-56, 1928. C. UHPUBLISH ED MATERIALS Gilbex*t, J. R., ’’The Traits of Secondary School Instrumentalists and Their Relationship to Achievement in Instrumental Music.” Unpublished M. A. thesis, Syracuse University, 1943. Gilpin, G. Noble, ” A Study Correlating Scores on KwalwasserDykema Tests of Musical Talent, Washburne Thaspic Personality Inventories, Otis Quick Scoring Mental Ability Tests, and an Original Phonophotographic Test.” Un­ published M. A. thesis, Syracuse University, 1941. Immel, Earle Barnard, ”An Experimental Investigation of the Relationship Between Musical Capacity and Emotional Status of High School Seniors.” Unpublished M. A. thesis, The University of Southern California, 1939. 166 Lamp, Charles, "The Determination of Aptitude for Specific Musical Instruments. " Unpublished Ph. D. thesis, University of California, 1935. Wagner, Doris, "A Comparison of Earned Scores by Junior High School Pupils In the Washburne Social Adjustment Inventory and the Kwalwasser-Dykema Tests of Musical Talent." Unpublished M, A. thesis, Syracuse University, 1946. •Yenaas, Sigurd B., "A Study of the Relation Between Musical Ability and Various Intelligence, Scholastic, and Personality Factors." Unpublished M. A. thesis, University of Idaho, 1940. D. PARTS OF SERIES Larson, William S., Measurements of Musical apt for the Predict^,an of Si^_geg-s. lil Instrumental Music. Psychological Monographs, Mo. 181, University of Iowa Studies in Psychology. Iowa City, Iowa: University of Iowa Press, 1930, Vol. 40, pp. 33-73. Mosher, K. M., A Study of the Group Method of Measurement of Sight-Singing. Contributions to Education, No. 194. New York: Teachers College, Columbia University, 192b, 75 pp. Saetveit, Joseph G., Carl E. Seashox*e, and Don Lewis, AgvA&I.oa o£ tji£ S&flfihfifg MgAgJageft o£ Musical Talents. Series on Alms and Progress of Research, No. 65. Iowa City, Iowat University of Iowa Press, 1940, 62 pp. E. PUBLICATIONS OF LEARNED ORGANIZATIONS Adkins, Dorothy C., Cons p n &£& Analysis Achievement Tests. Washington, D. C . : United States Government Printing Office, 1947, 292 pp. Manual of Instruction* ( _ _ _ ; _____ _______ Service, 1949, 4 pp. 1949 edition- Pi'ineeton. N e w _ t 166 Cooperative Reading ffppiPrff*renglQfl ?e£.tJBL: goasag.nAflg. New York: lp£j?.m&%XPJl interpretfttloji, sod. IZas.- Cooperative Test Service, 1940, 4 pp. F. TESTS American Council £ & Princeton, New Jersey: Educational Testing Service, Cooperative Test Division, 1949* Bernreuter, Robert O., JhS Personality Inventory. California: Stanford University Press, 1935. Cooperative Reading CaaacsfcgflgAsa IfiSJtiLCooperative Test Service, 1940. Stanford, New York: Kwalwasser, Jacob, and Peter Dykema, k » a !*tft«ser-Dvkema Music Tests. Nev: York: Carl Fischer, Inc., 1930. Seashore, Carl E., Measures of Musical Talent. Columbia Phonograph Co., 1919. New York: ________ . Measures o£ MdfiA