University Microfilms International tann Arbor, Michigan 48106

Transcription

1 EMIG, SANDRA JILL THE RELATIONSHIPS OF SELECTED MUSICAL, ACADEMIC, AND PERSONAL FACTORS TO PERFORMANCE IN THE FRESHMAN AND SOPHOMORE MUSIC THEORY AND EAR TRAINING SEQUENCES AT THE OHIO STATE UNIVERSITY. THE OHIO STATE UNIVERSITY, PH.D., 1978 University Microfilms International tann Arbor, Michigan Copyright by Sandra Jill Emlg 1978

2 THE RELATIONSHIPS OF SELECTED MUSICAL, ACADEMIC, AND PERSONAL FACTORS TO PERFORMANCE IN THE FRESHMAN AND SOPHOMORE MUSIC THEORY AND EAR TRAINING SEQUENCES AT THE OHIO STATE UNIVERSITY DISSERTATION Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Sandra Jill Emig, B.Mus., M.A. * * * * * The Ohio State University 1978 Reading Committee: Approved By A. Peter Costanza Burdette L. Green William Poland U/Jd School of Music Pcr&ujj

3 ACKNOWLEDGMENTS Grateful acknowledgment is made to my adviser, Professor William Poland, for his inspiration, encouragement, and guidance in the preparation of this dissertation. Gratitude is expressed to Professors A. Peter Costanza and Burdette Green for the constructive suggestions which they provided as readers and committee members. Appreciation is expressed to Professors George Policello and Ransom Whitney and Mr. Fred Ruland of The Ohio State University Department of Statistics for their assistance in the preparation and interpretation of the statistical data of this investigation. Finally, acknowledgment is made to The Ohio State University Instructional and Research Computer Center, Dr. Roy F. Reeves, Director, for the computer funding which made possible the execution of the statistical analyses. ii

4 VITA March 7, Born - Steubenville, Ohio B. Mus. in Euphonium Performance The Ohio State University Columbus, Ohio University Fellow The Ohio State University Columbus, Ohio Graduate Teaching Associate The Ohio State University Columbus, Ohio M. A. in Music Theory The Ohio State University Columbus, Ohio Instructor of Music Theory William Jewell College Liberty, Missouri PUBLICATIONS "The Musical Circles of Johann David Heinichen and Johann Mattheson", Journal of the Graduate Music Students at The Ohio State University 3 No. 6, Spring 1977, pp "Undergraduate Music Entrance Requirements and Policies of Ohio Colleges and Universities" with A. Peter Costanza, Triad, Vol. 45/2, November 1977, p. 50. ill

5 FIELDS OF STUDY Major Field: Music Studies in Music Theory. Professors Burdette Green, Norman Phelps, and William Poland Studies in Measurement and Evaluation. Professors A. Peter Costanza, Edwin Novak, and William Poland Studies in Music History. Professors Herbert Livingston, Martha Maas, Alexander Main, and Keith Mixter iv

6 TABLE OF CONTENTS Page ACKNOWLEDGMENTS... ii V I T A... ill LIST OF T A B L E S... x Chapter I. INTRODUCTION... *... 1 Purpose of the Investigation... 2 Definition of Terms... 3 Presentation of the Investigation II. REVIEW OF THE LITERATURE... 9 Salisbury and Smith (1929) 10 Wilson (1930) Tillson (1931) 13 Chadwick (1933) Farnsworth ( ) Taylor ( ) Poland (1960b) 19 White ( ) Roby ( ) Poland (1963) Perry ( ) v

7 Chapter Page Lee ( ) Davis ( ) Maier ( ) Melton (1973) Summary of the L i t e rature III. DESCRIPTIONS OF THE MUSIC THEORY AND EAR TRAINING SEQUENCES, THE VARIABLES, AND THE S U B J E C T S The Music Theory and Ear Training Sequences The Predictor Variables The Criterion Variables The S u b j e c t s IV. RESULTS OF THE INVESTIGATION Relationships between the Predictor and Criterion Variables Relationships between the Predictor Variables and the Music Theory Course Grade Criterion Variables Relationships between the Predictor Variables and the Ear Training Course Grade Criterion Variables Stepwise Multiple Regression Analyses Involving the Musi.a Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Criterion Variables vi

8 Chapter Page Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Music Theory Criterion Variables Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Ear Training Criterion Variables Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score Predictor Variables and the Criterion Variables.. 77 Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score Predictor Variables and the Music Theory Course Grade Criterion V a r i a b l e s Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score Predictor Variables and the Ear Training Course Grade Criterion V a r i a b l e s Stepwise Multiple Regression Analyses Involving the American College Tests Component Score Predictor Variables and the Criterion Variables Stepwise Multiple Regression Analyses Involving the American College Tests Component Score Predictor Variables and the Music Theory Course Grade Criterion V a r i a b l e s vii

9 Chapter Page Stepwise Multiple Regression Analyses Involving the American College Tests Component Score Predictor Variables and the Ear Training Course Grade Criterion V a r i a b l e s Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Criterion Variables 89 Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Music Theory Course Grade Criterion V a r i a b l e s Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Ear Training Course Grade Criterion V a r i a b l e s Application of the R e s u l t s Summary of the R e s u l t s V. SUMMARY OF THE INVESTIGATION AND RECOMMENDATIONS FOR FURTHER S T U D Y Restatement of the P r o b l e m Related Research... Ill The Music Theory and Ear Training Sequences, the Variables, and the Subjects Employed Results viii

10 Chapter Page Comparison of the Results from this Investigation and Results Found in Prior Investigations. 121 Recommendations for Further Study Discussion APPENDIX REFERENCES ix

11 LIST OF TABLES Table Page 1. Pearson Product-Moment Coefficients of Correlation Coefficients of Correlation (r) of Predictor Variables with the Music Theory Course Grade Criterion V a r i a b l e s Pearson Product-Moment Coefficients of Correlation Coefficients of Correlation (r) of Predictor Variables with the Ear Training Course Grade Criterion V a r i a b l e s Summary of the Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Music Theory Course Grade Criterion Variables Summary of the Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Ear Training Course Grade Criterion Variables Summary of the Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score Predictor Variables and the Music Theory Course Grade Criterion Variables Summary of the Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score Predictor Variables and the Ear Training Course Grade Criterion Variables Summary of the Stepwise Multiple Regression Analyses Involving the American College Tests Component Score Predictor Variables and the Music Theory Course Grade Criterion Variables x

12 Summary of the Stepwise Multiple Regression Analyses Involving the American College Teste Component Score Predictor Variables and the Ear Training Course Grade Criterion Variables... Summary of the Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Music Theory Course Grade Criterion Variables... Summary of the Stepwise Multiple Regression Analyses Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Ear Training Course Grade Criterion Variables... Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Music Theory Course Grade Criterion Variables Utilizing the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables... «... Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Ear Training Course Grade Criterion Variables Utilizing the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables... Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Music Theory Course Grade Criterion Variables Utilizing the Music Placement Test Battery Total Score Predictor Variables... Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Ear Training Course Grade Criterion Variables Utilizing the Music Placement Test Battery Total Score Predictor Variables...

13 Table 15. Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Music Theory Course Grade Criterion Variables Utilizing the American CoZZege Teste Component Score Predictor V a r i a b l e s Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Ear Training Course Grade Criterion Variables Utilizing the American CoZZege Tests Component Score Predictor Variables Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Music Theory Course Grade Criterion Variables Utilizing the Music PZacement Test Battery Total Score and American CoZZege Tests Component Score Predictor Variables Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Ear Training Course Grade Criterion Variables Utilizing the Music PZacement Test Battery Total Score and American CoZZege Tests Component Score Predictor Variables Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Music Theory Course Grade Criterion Variables Utilizing the Five Best Predictor Variables Regression Coefficients, Constants, and Standard Errors of Estimate for the Multiple Regression Equations of the Ear Training Course Grade Criterion Variables Utilizing the Five Best Predictor Variables Means, Standard Deviations, Minimums, and Maximums for All Predictor and Criterion Variables Pearson Product-Moment Coefficients of Correlation (r) for All Predictor and Criterion Variables.... Page xii

14 Table Page 23. Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the First Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Third Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Sixth Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratio for the Regression Analysis Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the First Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Third Quarter Ear Training Course Grade Criterion Variable xiii

15 Table Page 28. Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score, American College Tests Component Score, and Background Information Predictor Variables and the Sixth Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score Predictor Variables and the First Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score Predictor Variables and the Third Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score Predictor Variables and the Sixth Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score Predictor Variables and the First Quarter Ear Training Course Grade Criterion Variable xiv

16 Table Page 33. Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score Predictor Variables and the Third Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score Predictor Variables and the Sixth Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the American College Tests Component Score Predictor Variables and the First Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the American College Tests Component Score Predictor Variables and the Third Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the American College Tests Component Score Predictor Variables and the Sixth Quarter Music Theory Course Grade Criterion Variable xv

17 Table Page 38. Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the American College Tests Component Score Predictor Variables and the First Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the American College Tests Component Score Predictor Variables and the Third Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the American College Tests Component Score Predictor Variables and the Sixth Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music "Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the First Quarter Music Theory Course G-ade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Third Quarter Music Theory Course Grade Criterion Variable xvi

18 Table 43. Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Sixth Quarter Music Theory Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the First Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Analysis Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Third Quarter Ear Training Course Grade Criterion Variable Multiple Correlation Coefficients, Standard Error of Estimate, Constant, Regression Coefficients, Standard Errors of Regression Coefficients, and F Ratios for the Regression Avalysis Involving the Music Placement Test Battery Total Score and American College Tests Component Score Predictor Variables and the Sixth Quarter Ear Training Course Grade Criterion Variable... Page xv ii

19 CHAPTER I INTRODUCTION One of the requirements for obtaining any of the undergraduate degrees offered by The Ohio State University School of Music is the successful completion of the basic two-year sequences in music theory and ear training. These six-quarter sequences include study in the areas of music fundamentals, diatonic and chromatic harmony, sightsinging, dictation, and keyboard harmony (Infra, Chapter III, for descriptions of sequences).. The School of Music is vitally interested in predicting an applicant s performance in these sequences. Such knowledge greatly assists the School in admitting to the various undergraduate degree programs students who will, in all probability, successfully complete all of these courses. Also, such knowledge allows the School of Music to diagnose some of the major strengths and weaknesses in the music theory and ear training background of the applicant and to advise him or her of these strengths and weaknesses before beginning work in these sequences. The School of Music has been interested in the area of predicting student performance in music theory and ear training courses for nearly 50 years. The first person to report his research in this 1

20 area was Wilson (.1930), who determined the relationships between the Seashore Measures of Musical Talents, an original test battery, an original questionnaire, and the music theory and ear training grade point averages CInfra, pp , for description of Wilson study). In the 1950 s, Poland (1960c, 1963X developed The Ohio State University Music "Placement Test Battery to be used in the prediction of performance in music theory and ear training courses (Infra, pp.45-47, for description of test battery). Extensive research of the relationships between scores obtained on this test battery, The Ohio State University Psychological Examination, personal background information, and grades obtained in integrated theory and ear training courses was conducted (Infra, p. 24, for description of Poland study). The Music Placement Test Battery was subsequently included as part of the application process to the School of Music. Since the last investigation reported by Poland, the integrated music theory and ear training sequence has been divided into two separate entities that are taught and evaluated independently. The relationships reported, consequently, are not applicable specifically to the two sequences that are currently offered by the School of Music. Purpose of the Investigation In order to provide current predictive information for The Ohio State University School of Music concerning the performance of its undergraduate applicants in the freshman and sophomore music

21 theory and ear training courses, this investigation was designed to determine: 1. the relationships between individual predictor variables, representing musical achievement, general academic ability, and personal background information, and performance in freshman and sophomore music theory and ear training courses; 2. the relationships between selected groups of these predictor variables and performance in freshman and sophomore music theory and ear training courses; and 3. the differences which exist between the predictor variables that relate most highly to performance in music theory courses and those that relate most highly with performance in ear training courses. The results obtained in this investigation should be useful to The Ohio State University School of Music in predicting the performance of its applicants in music theory and ear training courses and in advising these applicants prior to beginning these courses. Definition of Terms Following are definitions of certain statistical and measurement terms which are included in the presentation of this investigation: 1. Criterion Variable: A criterion variable is a measure of success or failure to which other variables can be compared. Two

22 examples of criterion variables used in this investigation are the first quarter music theory course grade and the third quarter ear training course grade; 2. Predictor Variable: A predictor variable is a measure that is compared to the criterion variable. It is a measure that can be used to estimate the future performance of an individual on a particular task or routine. Two examples of predictor variables used in this investigation are the total score of the Music Recognition test of The Ohio State University Music Placement Test Battery and the number of years of private piano instruction; 3. Coefficient of Correlation (r): A coefficient of correlation is a measurement of the relationship between two variables. The coefficient of correlation can range from +1.0, which is a perfect positive correlation, through zero, which is complete independence, to -1.0, which, is a perfect negative correlation. This coefficient is not to be interpreted as a percentage, and so cannot be directly compared quantitatively with other correlation coefficients. The Pearson product-moment coefficient of correlation, symbolized by "r", is the most commonly computed and has been used throughout this investigation. To be consistent, all correlation coefficients presented in this investigation are rounded to two significant places; 4. Coefficient of Multiple Correlation (R): A coefficient of multiple correlation, symbolized by UR", is a measurement of the relationship between one variable and two or more different variables

23 taken together. The coefficient of multiple correlation is interpreted just as the coefficient of correlation. Again, all multiple correlation coefficients presented in this investigation are rounded to two significant places; 5. Coefficient of Determination (r2 or R2); The coefficient of determination is the proportion of variability of a criterion variable accounted for by the predictor variable(s). The percentage of this variability is computed by squaring the correlation coefficient or multiple correlation coefficient and multiplying the result by 100. This percentage is useful because, unlike the Pearson or multiple correlation coefficient, it can be directly compared quantitatively with other such percentages; 6. Standard Error: The standard error of a computed measurement defines limits around the computed value within which most of the cases lie. Sixty-seven (67) percent of the cases should lie within one standard error unit, 95 percent should lie within two standard error units, and 99 percent should lie within three standard error units on either side of the calculated measurement. For example, if a measurement were computed to be 100 with a standard error of 5, 67 percent of the cases should lie between 95 and 105, 95 percent should lie between 90 and 110, and 99 percent should lie between 85 and 115. All standard errors in this investigation are rounded to four significant places;

24 7. F Ratio: The F ratio is the ratio of obtained information to error in the computation of a multiple correlation coefficient. The F ratio of a multiple correlation coefficient is an index of the significance of the coefficient, while the F ratio of a predictor variable is an index of the significance of the increase made in the multiple correlation coefficient when that variable is included with other predictor variables. All F ratios in this investigation are rounded to two significant places; 8. Statistical Significance: Statistical significance is the level at which a determined relationship may be accepted as real and accounted for by factors other than chance. Throughout this investigation, all coefficients of correlation (r) and coefficients of multiple correlation (R) significant at the.01 level are underlined, indicating that there is a 99 percent chance that the obtained coefficient is significantly different from zero. F ratios significant at the.05 level are sumbolized by and those significant at the.01 level are symbolized by indicating that there is respectively a 95 and 99 percent chance that the correlation or difference tested is significantly different from zero; 9. Test Battery: A test battery is a set of tests designed to be administered together. The tests comprising the battery usually have been designed and developed as a unit, with the purpose of providing complete and efficient coverage of some unit of ability or achievement. Two examples of test batteries utilized in this

25 investigation are The Ohio State University Musie Placement Test Battery and the American College Tests; 10. Aptitude Test: An aptitude test is a test designed to measure what an individual can learn to do if he or she receives appro- propriate education or training. The example of an aptitude test battery utilized in this investigation is the American College Tests; 11. Achievement Test: An achievement test is a test designed to measure what an individual has learned to do as a result of direct or indirect experience, often that provided in school. The example of an achievement test battery utilized in this investigation is The Ohio State University Music Placement Test Battery; 12. Standardized Test: A standardized test is a test that has been designed and published for general use. The most distinctive feature of a standardized test or test battery is a set of norms based on some general reference population. Other usual features include selection of the items on the basis of preliminary tryout and analysis, standardized directions for administration and scoring, and a manual providing various types of statistical evidence about the test. The example of a standardized test battery utilized in this investigation is the American College Tests; and 13. Localized Test: A localized test is a test that has been designed for use at a particular institution. Such a test may or may not have a set of established norms, but usually does have standardized instructions for administration and scoring. The example of a

26 localized test battery utilized in this investigation is The Ohio State University Musia Plaoement Test Battery. 8 Presentation of the Investigation Chapter II will present a review of selected literature related to the prediction of student performance in the basic music theory and ear training courses required of all undergraduate music majors. Chapter III will present a description of the music theory and ear training sequences, the predictor and criterion variables, and the subjects employed in the investigation. Chapter IV will present the statistical and descriptive results of the investigation. Chapter V will present a summary of the investigation and recommendations for further study.

27 «CHAPTER II REVIEW OF THE LITERATURE Chapter I presented the purpose of this investigation and the definitions of terminology used in the report of the investigation. Chapter II will present a review of selected literature related to the prediction of student performance in the basic music theory and ear training courses required of all undergraduate music majors. Numerous research studies have been undertaken that relate to the prediction of student performance in the undergraduate basic music theory and ear training courses. In order to present those research studies most related to the research undertaken in this investigation, the following four criteria were established; 1. The criterion variables must measure performance in all or part of the required undergraduate courses in music theory and ear training. These measures may be course grades, grade point averages, or specialized examinations; 2. The subjects in the study must be the entire class of undergraduate music majors in a given year or set of years; 3. The research study must concern subjects at a single undergraduate school or college In the United States or Canada; and 9

28 10 4. Statistical data supporting the research study must be available. Reviews of the selected studies are presented below in chronological order. Salisbury and Smith (1929) One of the earliest studies in the field was designed to predict student performance in sightsinging courses at Bellingham (Washington) Normal School. Salisbury and Smith investigated the relationships between the scores obtained from three written tests as predictor variables and a subsequent sightsinging examination score as the cri- terior variable. Conducted over a three-year period, the predictor variables used in this study were the scores obtained from a locally constructed dictation test; the Pitch, Tonal Memory, Time, Rhythm, and Consonance tests of the Seashore Measures of Musical Talents, 1919 edition; and Part I of the Thorndike Examination for High School Graduates j series After making refinements in the localized dictation test and the sightsinging examination during the first two years of the study, Salisbury and Smith reported the findings from the third year of the study (p. 434), which are presented on the following page. As a result of their research, Salisbury and Smith concluded that the Pitch and Tonal Memory tests of the Seashore Measures of Musical Talents had "...very significant predictive value..." and

29 11 Predictor Variable Sightsinging Examination Score Dictation Test 79 r Seashore Measures of Musical Talents Consonance Pitch Rhythm Time Tonal Memory Thorndike Examination for High School Graduates Part I..27 N=144 that "the other... Seashore tests were of decidely less value..." (p. 438). In addition to the above cofficients of correlation, a multiple correlation coefficient (R) of.84 was calculated using the localized dictation test and the Seashore Pitch and Tonal Memory tests as predictor variables. The inclusion of any fourth variable in the multiple regression analysis did not effect a significant increase in the multiple coefficient of correlation (p. 435). Wilson (1930) Wilson sought to determine the degree of correlation between various test scores and background factors and subsequent performance in a music major program at The Ohio State University. His predictor

30 variables included scores from both standardized and localized tests. His standardized tests were all six tests of the Seashore Measures of Musical Talents edition, while his localized test battery comprised tests of Tonic Memory, Resolution, and Score Reading. A background questionnaire was used to gather information concerning the students' home situation, music lessons, music ideals, and music experience. Among the criterion variables obtained were the grade point averages (GPA) for the first year of study in music theory and ear training courses. The subjects for the study were the 83 freshmen music majors who entered the music department at The Ohio State University in the autumn of Wilson reported the correlation coefficients from the study (p. 87), which are presented on the following page. As a result of his findings, Wilson concluded that his localized tests "...[had] a greater validity than the prognostic tests based on physiological limitations, which [had] chiefly been used up to the present time" (p. 100). In fact, however, only a portion of Wilson's localized predictor variables correlated higher with the criterion variables than did scores from the Seashore Measures of Musical Talents.

31 Predictor Variable Theory GPA Ear Training GPA 13 r r Seashore Measures of Musical Talents Consonance Intensity Pitch Rhythm Time Tonal Memory Total Wilson Music Test Battery Resolution Score Reading Tonic Memory Total Wilson Questionnaire Home Situation Music Lessons.29^.07 Music Ideals.37^.39 Music Experience Total N=83 Tillson (1931) Tillson investigated the relationships between test scores obtained from the Seashore Measures of Musical Talents* 1919 edition, the Kwalwasser-Dykema Music Tests* the American Council on Education Psychological Examination* 1929 edition, and subsequent performance in music ear training courses by students enrolled at Indiana State Teachers College from 1926 through In his study, he used the grade point ayerage (GPA) in ear training courses as the criterion variable.

32 14 Tillson found the following coefficients of correlation (pp ; 119): Predictor Variable N Ear Training GPA American Council on Education Psychological Examination r Kwalwasser-Dykema Music Tests Intensity Discrimination Melodic Tasta Pitch Discrimination Pitch Imagery Rhythm Discrimination Rhythm Imagery Time Discrimination Tonal Memory Tonal Moyement Tonal Quality Discrimination Seashore Measures of Musical Talents Consonance Intensity Pitch Rhythm Time Tonal Memory Concerning his results, Tillson concluded (p. 125): It would seem that term grades in ear training and sightsinging are affected about equally by musical talent and general mental powers. Since this is true, both musical talent and intelligence of the student should be taken into consideration in any attempt to decide which students should be permitted to take training in courses leading to licenses in music supervision. However, it should be noted that the Seashore Tonal Memory test alone

33 accounted for more than twice the variability of the criterion variable than did the American Council an Education Psychological Examination. 15 Chadwick (1933) This study sought to determine the relationships between scores obtained from various standardized academic and music tests and subsequent performance in sightsinging courses at Colorado State Teachers College. The data collected concerning the 39 music students in the investigation were the sightsinging grade point average (GPA); a composite score from the Tonal Memory, Rhythm, Time, Intensity, and Pitch tests of the Seashore Measures of Musical Talents, 1919 edition; and the raw scores from each the Teachers College Achievement Test and the American Council on Education Psychological Examination edition. Chadwick reported the following coefficients of correlation (p. 672): Predictor Variable Sightsinging GPA r Seashore Measures of Musical Talents Teachers College Achievement Test American Council on Education Psychological Examination.64 N=39

34 16 Furthermore, a multiple coefficient of correlation (R) of.84 between the test scores and the sightsinging grade point average was obtained in this study (p. 674). Farnsworth (1935) This study was designed to determine whether standardized tests of musical capacity or of intelligence could better predict several types of music grades, including those of music theory and ear training. Conducted at San Jose (California) State Teachers College, the music capacity tests used were the Pitch and Tonal Memory tests of the Seashore Measures of Musical Talentss 1919 edition. The intelligence tests used were the American Council on Education Psychological Exami- nationj 1929 edition, and the Iowa High School Content Examination. The criterion variable employed in the study was the grade point average (GPA) of the music theory and ear training courses. Subjects for the study were 96 students enrolled in the basic music theory and ear training courses at the College. Farnsworth reported the coefficients of correlation from the study (p. 349), which are presented on the following page. In addition, Farnsworth computed a multiple correlation coefficient (R) of.28 between the two tests of music capacity and the music theory and ear training grade point average, while that between the two intelligence tests and the grade point average was computed to be.27. The multiple correlation coefficient (R) between all four tests and the criterion variable was.38.

35 17 Predictor Variable Music Theory and Ear Training GPA r Music Capacity Tests Seashore Sense of Pitch Seashore Tonal Memory Intelligence Tests American Council on Education Psychological Examination Iowa High School Content Examination N=96 Farnsworth summarized these findings by stating that 1...neither type of test appeared to be significantly superior in the prediction of the... grades from music theory" (p. 350). Taylor (1941) This study was conducted at the College of Music of Cincinnati (Ohio) in an attempt to evaluate a battery of musical and academic tests as a basis for prognosis of success in a college of music. A five-year testing program at the College was conducted from 1930 to 1935, employing the Seashore 'Measures of Musical Talents, 1919 edition; the Ewalwasser-Dykma Music Tests; the Melodic and Harmonic Sensitivity tests of the Ewalwasser Music Tests; the Measures of Musical Background* an original test devised by C. H. Taylor for experimental purposes at the college; and Forms V and W of the Detroit Advanced Intelligence Test, 1924 edition. The scores obtained on these tests

36 18 were then correlated with course grade point averages (GPA) in undergraduate harmony, sightsinging, and dictation. Taylor determined the following coefficients of correlation (pp ): Predictor Variable Harmony GPA Sightsinging GPA Dictatior GPA N r N r N r Detroit Intelligence Test Kwalwasser Music Tests Harmonic Sensitivity Melodic Sensitivity Kwalwasser-Dykema Music Tests Intensity Discrimination Melodic Taste Pitch Discrimination Pitch Imagery Rhythm Discrimination Rhythm Imagery Time Discrimination Tonal Memory Tonal Movement Tonal Quality Discrimination Measures of Musical Background Background Discrimination of Mode Background Discrimination of Rhythm Seashore Measures of Musical Talents Consonance Intensity Pitch Rhythm Time Tonal Memory

37 Taylor concluded these results with the following conjecture 19 (p. 14):... A combination of tests (Background Discrimination of Mode, Pitch Imagery, and Tonal MemoryX with an intelligence test {Detroit Advanced Intelligence Test) may offer a prediction sufficiently accurate to be practical. Poland (1960b) This study Was designed in part to determine the relationships existing between performance on certain variables, representing elemental parts of the aural and notational language of music, and success in music theory and ear training at certain criterion points in the basic two-year college-level courses in music theory and ear training. The investigator-constructed musical achievement test was designed to investigate performance in three major categories of the fundamentals of music, namely chords, intervals, and scales. Four variables were measured in each, of these categories. These variables were notation accuracy, defined as the ratio of the number of notation items correct to the number of notation items tried in the allotted time; notation output, defined as the ratio of the number of notation items correct in the allotted time to the number of notation items presented; notation speed, defined as the ratio of the number of notation items attempted to the number of notation items presented; and aural identification, defined as the ratio of the number of aural items correct to the number of aural items presented. One of the groups of students to which

38 this test was administered was one just beginning the first music theory and ear training course at The Ohio State University. The two criterion variables for this group of students were the first quarter music theory and ear training course grade and the first year music theory and ear training grade point average (GPA). Poland reported the following correlation coefficients between the 12 predictor variables and the criterion variables (pp. 105; 107): Predictor Variable First Quarter First Year Music Theory and Music Theory and Ear Training Grade Ear Training GPA Chord Notation Accuracy Notation Output Notation Speed Aural Identification Interval Notation Accuracy Notation Output Notation Speed Aural Identification Scale Notation Accuracy Notation Output Notation Speed Aural Identification r r N=79 N=58 In addition, the multiple correlation coefficient ( ) between all 12 predictor variables and the first quarter music theory and ear training grade was computed to be.66 (p. 105)., while that between all

39 12 predictor variables and the first year music theory and ear training grade point average was computed to be.80 (p. 107). 21 White (1961) This study was designed to determine whether a test of musical achievement or one of musical aptitude could better predict performance in freshman and sophomore music theory and ear training courses at St. Olaf (Minnesota) College. The musical achievement and aptitude tests that were used in this study were respectively the Aliferis Music Achievement Test and the Pitch, Loudness, Rhythm, and Tonal Memory tests of the Seashore Measures of Musical Talents revision. Performance on these two tests were then correlated with the grade point average (GPA) of four semesters of music theory and ear training study. Subjects for this study were the 59 music students who graduated from the College in 1960 and White reported the results from his investigation (pp ), which are presented on the following page. He then concluded that the Aliferis Music Achievement Test was "...far superior to the Seashore Measures of Musical Talents as a predictor of success for the 59 students studied" (p. 316). When used with other available measures, White found the Aliferis Music Achievement Test to be an important tool in the predicting of success in music theory and ear training courses at St. Olaf College. These other measures included a piano test, the American

40 22 Predictor Variable Music Theory and Ear Training GPA Aliferis Music Achievement Test.63 r Seashore Measures of Musical Talents Loudness Pitch Rhythm Tonal Memory N=59 Council on Education Psychological Examination, and rank in the high school graduating class. The Seashore Measures of Musical Talents, on the other hand, offered little help in this prediction process (pp ). Roby (1962) This study was conducted at the University of Minnesota to determine the prognostic abilities of the scores obtained from the 1956 revision of the Seashore Measures of Musical Talents, the Aliferis Music Achievement Test, the American Council on Education Psychological Examination, and the University of Minnesota English Entrance Test. The criterion variable in the study was the grade point average (GPA) for all music theory and ear training courses in the first two years of music study. The subjects for this study were music students entering the University from 1953 through 1955 who completed the required two years of music theory and ear training study.

41 23 (p. 141) : Roby reported the following coefficients of correlation Predictor Variable N Music Theory and Ear Training GPA Aliferis Music Achievement Test Harmony Melody Rhythm Total Total Minus Rhythm Seashore Measures of Musical Talents Loudness Pitch Rhythm Timbre Time Tonal Memory Total American Council on Education Psychological Examination University of Minnesota English Entrance Test Roby also reported, although without statistical evidence, that student grades in music theory courses rarely deviated from the first year of study to the second, or that "... students who received A's the first year tended to continue to get A's, B students to get B*s, and so on" (p. 139).

42 Poland (1963) Poland conducted an extensive investigation of the prognosis of success in an undergraduate music major program at The Ohio State University School of Music (See also Poland, 1956; 1957; 1960a; 1960c). The predictor variables collected concerning each applicant to the School of Music were scores from The Ohio State University Music 'Placement Test Battery^ the total percentile from The Ohio State University Psychological Examination3 sex, age, major performance area, number of years of private study in the major performance area, number of years of private piano study, and participation in high school performance groups, music theory courses, and music history or appreciation courses. The criterion variables used throughout the study were grades and grade point averages of all course work completed at the University. In a multiple regression analysis of the predictor variables with first quarter music theory and ear training grades, Poland determined the multiple correlation coefficient (R) to be.83 (N=91). When a correlation study was made of the predicted grades of the first quarter course in music theory and ear training with the obtained grades in this course, a Pearson product-moment correlation coefficient (r) of.67 (N=87) was obtained (Data Supplemnt, p. 6). Perry (1965) This study was an analysis of the capability of scores on selected tests for predicting music theory and ear training proficiency

43 25 scores at the college freshman level. The tests Perry selected for use in this study were Forms A and B of the Drake Musical Aptitude Tests edition; the North Texas State University Freshman Placement Theory Examination edition; Form Ba of the Gordon Index of Music Insight edition; the Kxalwasser-Dykema Music Tests; the Gamma Test of the Otis Quick-Scoring Mental Abilities Tests3 Form AM; and the Wing Standardised Tests of Musical Intelligence edition. The criterion variables used were standard scores on tests developed for the investigation, in the areas of rhythmic dictation, melodic dictation, harmonic dictation, sightsinging, part writing, keyboard recognition and harmony, and music fundamentals. Subjects for the study were 91 music majors who completed their first year of music study at North Texas State University during the academic year. The coefficients of correlation from the Perry study (p. 131) are presented on the following page. In addition, Perry computed a multiple regression analysis, with the predictor variables being the rhythm score of the Drake Musical Aptitude Tests and the total scores from the Freshman Placement Theory Examination3 four selected tests of the Fwatwa&ser-Dykema Music Tests3 the Wing Standardized Tests of Musical Intelligence3 the Gordon Index of Music Insight3 and the Otis Quick-Scoring Mental Abilities Test. The criterion variable was the standard score of the total proficiency test. Using these six predictor variables, the multiple correlation coefficient (R) was.88 (p. 135).