THE RELATIONSHIPS OF PARENTAL INVOLVEMENT, MOTIVATING FACTORS, AND SOCIOECONOMIC STATUS TO HIGH SCHOOL ALL-STATE CHOIR AND BAND MEMBERSHIP Except where reference is made to the work of others, the work described in this dissertation is my own or was done in collaboration with my advisory committee. This dissertation does not include proprietary or classified information. ______________________________ Stephen Clyde Hickok Certificate of approval: ______________________________ ______________________________ Jane M. Kuehne Kimberly C. Walls, Chair Assistant Professor Professor Curriculum and Teaching Curriculum and Teaching ______________________________ ______________________________ Thomas R. Smith David M. Shannon Professor Professor Music Educational Foundations, Leadership, and Technology ______________________________ George T. Flowers Dean Graduate School THE RELATIONSHIPS OF PARENTAL INVOLVEMENT, MOTIVATING FACTORS, AND SOCIOECONOMIC STATUS TO HIGH SCHOOL ALL-STATE CHOIR AND BAND MEMBERSHIP Stephen Clyde Hickok A Dissertation Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Auburn, Alabama May 9, 2009 iii THE RELATIONSHIPS OF PARENTAL INVOLVEMENT, MOTIVATING FACTORS, AND SOCIOECONOMIC STATUS TO HIGH SCHOOL ALL-STATE CHOIR AND BAND MEMBERSHIP Stephen Clyde Hickok Permission is granted to Auburn University to make copies of this dissertation at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. __________________________ Signature of Author __________________________ Date of Graduation iv VITA Stephen Clyde Hickok was born to Fred and Marguerite Hickok in Danville, Pennsylvania, on December 28, 1954. He has two brothers, Fred Hickok and Philip Hickok, and two sisters, Deborah Hickok and Naomi McChesney. Mr. Hickok attended Asbury College in Wilmore, Kentucky and Westminster College in New Wilmington, Pennsylvania where he earned his Bachelor of Music degree in Church Music in 1976. He attended Asbury Theological Seminary and then University of Kentucky where he earned his Masters of Music degree in Vocal Performance in 1981. Mr. Hickok served as the Minister of Music in churches in Kentucky, Florida, Georgia, and Alabama. He was an Assistant Professor of Music and Director of Choral Activities at Andrew College in Cuthbert, Georgia. Mr. Hickok married Karen Regina Stechman, daughter of William and Elaine Stechman, in 1979. They have a daughter Melissa, born April 23, 1982 and a son William, born December 30, 1987. v DISSERTATION ABSTRACT THE RELATIONSHIPS OF PARENTAL INVOLVEMENT, MOTIVATING FACTORS, AND SOCIOECONOMIC STATUS TO HIGH SCHOOL ALL-STATE CHOIR AND BAND MEMBERSHIP Stephen Clyde Hickok Doctor of Philosophy, May 9, 2009 (M.M., University of Kentucky, 1981) (B.M., Westminster College, 1976) 153 Typed Pages Directed by Kimberly C. Walls The primary purpose of this study was to identify differences in parental involvement, motivational factors, and socioeconomic status (SES) between high school band and chorus participants. The secondary purpose of this study was to investigate the relationship of these factors to students who participate in all-state music festivals. The participants (N = 403) in this study included choral and band students from six high schools in the southeastern United States. Participants completed the Characteristics of High School Music Students Survey (CHSMSS) which measured vi aspects of students? home environment, parental support in music, and students? attributions of success in music. Performance achievement was operationalized as students? participation in all-state chorus or band festivals. SES was measured through Hollingshead?s Two-Factor Index of Social Position. Nine factors related to band, chorus, and all-state participation were found. Discriminant Function Analysis indicated a significant difference between band and chorus students in SES and musical ability attribution. All factors related to family environment and parental involvement were significantly higher for all-state participants than for non-all-state participants. Parental involvement was a significant factor in students? performance achievement in band and chorus. vii ACKOWLEDGMENTS The author would like to thank Dr. Kimberly Walls, Dr. Thomas Smith, Dr. David Shannon, and Dr. Jane Kuehne for their assistance throughout my doctoral program and with my research. I have been fortunate to have an advisory committee that has provided exceptional support and encouragement throughout my doctoral studies. Each member of my committee has been generous with their time and constant in their positive attitude. I am thankful to Dr. Walls for her practical and insightful teachings in music education and her guidance in my research. She has been helpful, thorough, encouraging, and always available as an advisor and mentor. I am thankful to Dr. Smith for his guidance in my music studies and his encouragement and insights in my research endeavors. I am thankful to Dr. David Shannon for his guidance and expertise in the methodology and statistics of my research. I am thankful to Dr. Kuehne for the insights and ideas she has provided me in my research. Most importantly, I am grateful to my wife, Karen, for her perseverance and support. My doctoral studies have been an extra burden for her. She has taken on additional burdens of financial support for the family during this time and has supported me in my endeavor. She has often repeated the motivating words that my father has used during the past few years; ?Is it done yet?? viii Style manual or journal used: Publication Manual of the American Psychological Association, Fifth Edition Computer software used: Microsoft Word 2007, Microsoft Excel 2007, Statistical Package for the Social Sciences (SPSS) 16.0, Analysis of Moment Structures (AMOS) Graphics 16.0 ix TABLE OF CONTENTS LIST OF TABLES ...................................................................................................... xii I. INTRODUCTION ................................................................................................1 Defining Success ...................................................................................................2 Measurements of Success .....................................................................................6 Research on Instrumental and General Music Students .....................................10 Purpose of the Study ...........................................................................................11 Research Questions .............................................................................................11 Assumptions ........................................................................................................11 Delimitations .......................................................................................................12 Definitions...........................................................................................................12 II. LITERATURE REVIEW ...................................................................................14 Overview of Research into Predictors of Musical Achievement ........................15 Music achievement, intelligence, and musical aptitude ......................................16 Family Background .............................................................................................22 Socioeconomic Status .........................................................................................31 Socioeconomic measurement instruments ..............................................33 Motivation and Attribution Theory .....................................................................37 x All-State Participants ..........................................................................................45 Summary .............................................................................................................47 III. METHODS AND PROCEDURES.....................................................................48 Participants ..........................................................................................................48 All-State Selection Criteria .................................................................................48 School Selection Procedures ...............................................................................51 Access and Permission ........................................................................................52 Data Collection Procedures .................................................................................53 Instrument ...........................................................................................................54 Evaluation of Pilot Study ........................................................................54 Structure of the CHSMSS .......................................................................57 Data Analysis ......................................................................................................59 IV. RESULTS ...........................................................................................................63 Introduction .........................................................................................................63 Response Rate and Sample Characteristics ........................................................63 Research Question 1 ...........................................................................................65 Reliability ................................................................................................65 Parental Involvement ..............................................................................67 Analysis of Motivating Factors ...............................................................67 Factors Related to Band, Chorus, and All-state Participation ............................73 Research Question 2 ...........................................................................................74 Research Question 3 ...........................................................................................76 xi Correlations of Variables in the PIM ..................................................................78 V. DISCUSSION .....................................................................................................83 Response to Research Questions ........................................................................83 Predicting Success in Band and Chorus..............................................................85 Implications for Parents ......................................................................................85 Implications for Educators ..................................................................................86 Recommendations for Further Research .............................................................87 REFERENCES .............................................................................................................89 APPENDIX A: Characteristics of High School Music Students Survey ...................102 APPENDIX B: Characteristics of High School Music Students Survey: Pilot Version .....................................................................................107 APPENDIX C: Office of Human Subjects Approval ................................................113 APPENDIX D: Communication and Permission Letters ...........................................124 APPENDIX E: Rotated Component Matrix for 35 AMF Variables .........................133 APPENDIX F: Rotated Component Matrix for 34 AMF Variables .........................136 APPENDIX G: Rotated Component Matrix for 34 AMF Variables Suppressed to Five Factors ...............................................................139 xii LIST OF TABLES 1. Hollingshead?s occupational and educational scale .........................................37 2. Vocal associations audition requirements ........................................................50 3. SES of pilot study participants .........................................................................56 4. Summary of participants by state and ensemble type ......................................65 5. Reliability of Zdzinski?s PIM scale and sub-scales ..........................................66 6. Reliability of Asmus? AMF scale .....................................................................67 7. Fit indices of the Asmus model and study data ................................................69 8. Percentages of variance explained in the 34 item AMF scale ..........................72 9. Descriptive statistics for CHSMSS factor means .............................................74 10. DFA means for band and chorus participation .................................................76 11. DFA means for all-state and non-all-state participation ...................................78 12. Pearson Product-Moment Correlation for variables in the PIM .......................79 1 CHAPTER 1 INTRODUCTION Why do some students achieve high levels of success in band and others do not? Why do some students achieve high levels of success in chorus and others do not? Would the student who is successful in band be as successful in choir? Are we placing students in band or chorus because of their best opportunity for musical achievement or is their placement a matter of luck? All humans are assumed to possess a capacity for musical competence but not all students exhibit the same level of musical competence in the music classroom (Hallam & Shaw, 2002). Music educators have continually sought to implement teaching strategies that contribute to students? competence and achievement in music. When music teachers can identify factors that lead to students? achievement in music, the music teachers should be better prepared to implement teaching strategies that help each student reach a higher level of achievement in music. Parents who understand the factors related to music achievement should be better prepared to implement strategies within a child?s environment that will enhance the child?s likelihood of high achievement in music. Researchers in the disciplines of music and psychology have identified parental support, teacher characteristics, academic success, music aptitude, music background, achievement motivation, student attributions of success, and personality 2 as factors related to student music achievement. However, at least three problems remain in predicting student achievement in music. The first problem in predicting music achievement is that researchers and educators have not reached a common definition of music achievement and what constitutes success in that achievement. The second problem is that music is too multifaceted to assess in any one test (Hufstader, 1974). Comprehensive tests over all possible facets of musical ability would be unwieldy to administer. The third problem is that although researchers have investigated success in instrumental students and general music students, few researchers have studied music achievement in vocal students. Defining Success One of the difficulties in identifying predictors of musical success is a lack of consistency in the criteria that constitutes musical competence (Boyle, 1992, p. 247). The definition of musical success varies among researchers, musicians, and teachers. To some, the quality of a student?s musical performance demonstrates success; to others, the development of a student?s musical abilities is the definition of success (Reimer, 2003, p. 48). Nadia Boulanger contends that the essence of music is the final product (Reimer, 2003, p. 48). Elliott (1995, p. 39-40) defines music as a process. The lack of consistency in the definition of musical aptitude, ability, and achievement among researchers has led to confusion in the measurement of musical success. This confusion has resulted in an unwillingness among teachers to rely on research in their teaching practices (Boyle, 1992, p. 247). 3 Researchers in music must identify whether they are measuring success according to aptitude, achievement, or ability. Performance assessment identifies a specific form of music ability. It is not synonymous with music aptitude (Gordon, 1998, p. 17). The constructs that are used to measure music ability are not the same constructs that are used to measure music aptitude. Hallam (1998) suggested that the needs for aural, cognitive, technical, musicianship, performance, and learning skills differ according to the type of music that is being performed. Hallam noted that jazz performance skills and orchestral performance skills are not the same. Hallam and Shaw (2002) suggested a definition of musical ability in their qualitative study of teachers and students over the age of 14. Their sample included children and teachers in a city school, a music school, and an institution of higher education. The researchers divided the participants (N = 490) into eight groups. The groups included musicians (n = 55), educators in non-music subject areas (n = 80), adults not involved in music (n = 47), adults who were moderately proficient on an instrument (n = 20), adults who were minimally proficient on an instrument (n = 106), students with more than 2 years of musical experience (n = 135), students with less than 2 years of musical experience (n = 33), and students with no musical experience (n = 14). The participants completed a one-question survey that began with the phrase ?Musical ability is . . .? The major concepts of musical ability that were identified from the composite of all the participants in Hallam and Shaw?s study were rhythmic ability, organization of sound, emotional sensitivity to music, communication, motivation, personal characteristics, and a combination of complex skills and ensemble skills. 4 Other researchers have used music teacher assessments as indicators of student success in music. In Bonifati?s (1997) study, instrumental music teachers defined student success as possessing good musical technique, possessing self discipline, actively participating in class, having a good attitude, enjoying music, developing a life-long love of music, and being a productive member of a performing group. The teachers characterized successful students as those who made steady progress, gained an appreciation of music, ably interpreted music, and developed a refined approach to creating art. Helwig and Thomas (1973) operationalized students? success through teachers? grades that were based on attitude, effort, and musicality. The assessment of students? success was subjective in nature, but Helwig and Thomas established an internal consistency of this assessment (r - .460, p ? .05) through the Spearman-Brown prophecy formula. Hufstader (1974) used the recommendations of band directors to identify successful and unsuccessful instrumental students. Each band director classified the students as low achievers, medium achievers, and high achievers. Hufstader used only the low group and the high group in order to establish two distinct achievement groups. One of the students identified by a teacher as a low achiever scored consistently with the high achievers on the tests. It is not clear whether this inconsistency was attributable to variances in the factors predicting success or to the teacher?s improper classification of the student?s success. Additional difficulties are associated with reliance on teacher assessment to operationalize student success in music. Pascoe and Waugh (2001) recognized the need to assess music students in the schools in Western Australia. They found that no 5 assessment methods or standards for assessments existed to objectively identify a student?s success in music. Pascoe and Waugh outlined five concerns in assessment. Assessments based on phrases such as ?participates enthusiastically,? ?enjoys music,? or ?practices regularly? did not indicate skills or abilities in music. Teachers had no objective measures for assessment. No benchmarks existed for teachers to monitor student progress; therefore, students were not assessed at periodic intervals. Without proper assessment of the students, teachers could not be held accountable for the achievement of the students. Researchers and educators have used varying levels of objectivity to interpret measurements of success. Teacher recommendation and expert judging in performance are the most commonly used methods of identifying musical success in high school music programs. Teacher recommendation is subject to the assessment measures of each teacher. Establishing reliability from the assessment of individual teachers is difficult (Helwig & Thomas, 1973; Hufstader, 1974). When multiple judges assess a single performance or achievement, inter-judge reliability can be established by comparing judges? scores. For example, Young (1971) utilized expert judges to determine success in beginning instrumental students. In Young?s study, three judges evaluated the recorded student performances. The judges scored the students independently. Young established reliability through inter-reliability coefficients among each pair of judges (r = .89, r = .75, r = .73) as well as for all the judges combined (r = .98). 6 The use of expert judges typically takes place only when a student participates in a musical performance outside the local school environment. If a student does not participate in solo and ensemble festivals or all-state competitions, it is unlikely that the student will be individually assessed by expert judges. The evaluation of students? success in music by a teacher or by expert judges is subject to the teacher?s understanding of success or the criteria for success used by the judges. One of the most commonly used measurements of success in public school band and choral programs is the judging system for all-state participation. All-state is a performance- based festival; therefore, adjudication of all-state students utilizes performance-based criteria. Students who are accepted for participation in all-state festivals have been assessed as successful in performance achievement. Measurements of Success Researchers have developed measurements based on an array of criteria to identify success in music. Tests of musical ability that were developed by Revesz, Seashore, Wing, and Gordon in the early and mid-twentieth century were built around the aural perception of music (Hallam & Shaw, 2002). Researchers have frequently identified musical aptitude and academic ability as two of the strongest predictors of musical success (Kuhlman, 2005). Kuhlman found that research on the effect of musical aptitude and academic ability presented inconsistent and often opposing results. Gordon (1986) reported that musical aptitude accounted for 37% of the variance of success in music achievement for beginning instrumentalists. According to Gordon (1967), academic achievement contributed little beyond musical aptitude to 7 enhance musical performance achievement. Klinedinst (1991) and McCarthy (1974), however, identified academic achievement as the most significant factor in instrumental music achievement. Success in instrumental performance significantly correlated to academic success when the performance criteria were based on music reading and notation skills. Reading and notation skills are academic in nature (Klinedinst, 1991). McCarthy (1974) described both music performance and academic achievement as academic activities. Four of the criticisms of educational testing that are pertinent to testing for musical achievement or ability are selection of appropriate tests, erroneous and na?ve interpretations of test results, confusion between the use of norm- and criterion-based testing, and extraneous variables influencing the test scores (Boyle & Radocy, 1986, p. 21-26). Researchers in music must be clear as to what aspect of music they are attempting to identify. Many students might possess a high aptitude for music that goes unidentified in the assessment procedures used within a chorus or band class. Students who are assessed according to the performance abilities they demonstrate are often identified as having musical aptitude. Performance assessment identifies a specific domain of music ability. It does not imply an assessment of music aptitude (Gordon, 1998, p. 3). The constructs that are used to measure music ability are not the same constructs that are used to measure music aptitude. Reliability is a concern when developing testing instruments, because the test must appropriately measure the aspect of musical success that the researcher intents to measure. Boyle (1992, p. 145-146) pointed out that Wing?s approach to assessing 8 students in music was to evaluate musical intelligence. Seashore?s approach focused more on the evaluation of musical ability (Boyle, 1992, p. 143-145). Gardner?s approach to the evaluation of musical intelligence varied from Wing?s approach in that Gardner isolated musical intelligence from other aspects of IQ. Music ability, music aptitude, music intelligence, music capacity, music talent, music sensitivity, musicality, and music achievement are terms that have been used to identify success in music (Boyle, 1992, chap. 16). Because of the multifaceted nature of musical ability and the varying definitions of success, a broad array of measurements is needed to predict success in music. Manor (1950) recommended the use of a broad array of psychometric measurements to predict success. Manor measured music aptitude, IQ, persistence, and tonette class achievement in directing fourth-grade students toward the appropriate instrumental studies. The Manor Persistence Ranking Scale used to rate the students? success in the tonette classes measured tone production, range, fingering, physical execution, tone quality, and interest in the instrument. Hallam and Shaw (2002) asked a group of respondents (n = 490) to rate the significance of 19 items to musical ability. The respondents included a spectrum of individuals from accomplished musicians to non-musicians. The ages of the respondents varied, but all were 14 years or older. The items included knowledge of music, music reading, composition and improvisation, evaluation of music, technical skills, appreciation of music, creativity, motivation, communication, and a variety of musical skills. The respondents viewed the nature of musical ability differently according to the extent to 9 which the respondents were actively involved in the process of making music. The study was inconclusive concerning the definition of musical success, but respondents overall viewed musical ability as learned rather than innate. Rainbow (1965) investigated 14 variables? association with musical aptitude. Rainbow described music aptitude as the potential talent that a student has for music and clearly differentiated between musical aptitude and musical achievement. Music achievement was one of the variables that Rainbow used to identify music aptitude. The variables used in the study were 1) pitch discrimination, 2) tonal memory, 3) rhythm, 4) musical memory, 5) IQ, 6) school achievement, 7) sex, 8) chronological age, 9) musical achievement, 10) musical training, 11) home enrichment, 12) interest in music, 13) participation in music by relatives, and 14) socioeconomic background. Correlations differed among elementary (n = 91), junior high (n = 112), and high school (n = 88) students. A multiple regression analysis revealed that the variables that contributed most to the variance in musical aptitude among all students were tonal memory (? = 2.93), IQ (?= 2.21), music achievement (? = 15.77), home enrichment (?= 3.72), interest in music (? = 4.06), and socioeconomic background (? = 4.05). Rainbow concluded that the three extra-musical variables that best predicted music aptitude were interest in music, home enrichment, and socioeconomic background. No single test of musical aptitude or musical intelligence has adequately predicted success in music (Boyle & Radocy, 1986; Hufstader, 1974). Even combinations of tests used in studies by Gordon (1967) and Young (1971) could not fully account for students? success in all aspects of music. 10 Research on Instrumental and General Music Students Research into factors contributing to musical success has often focused on instrumental students (Bonifati, 1997; Davidson, Howe, Moore, & Sloboda, 1996; Doan, 1973; Fitzpatrick, 2006; Hufstader, 1974; Klinedinst, 1991; Manor, 1950; McCarthy 1974; Pitts, Davidson, & McPherson, 2000; Schmidt, 2005, Sloboda & Howe, 1991; Young, 1971; Zdzinski 1992, 1993, 1996). Research on instrumental students has been directed toward children beyond the fourth grade. The Seashore, Wing, and Gordon tests were designed for students in grades 4 - 12. Most of the additional research found concerning success in other areas of music has also been directed toward 4th- through 12th-grade students. Gordon focused on that age group because he believed that children?s musical aptitude did not become stable until they reached the fourth grade (Gordon, 1998, p. 50, 63-64). Several researchers have identified factors that are related to student success in music by combining instrumental students and choral students into one study group (Asmus, 1985b, 1986a, 1986b; Brand, 1985, 1986; Br?ndstr?m, 2000; Dunlap, 1975; Greenberg 1970; Hallam & Shaw, 2002; Legette, 1998, 2003). Helwig and Thomas (1973) suggest that the reason for a lack of research on vocal students is that assessing vocal progress is more difficult than assessing instrumental progress. They note that tests designed to measure technical skills and facility of instrumental students provide for an objective assessment of the students. The progress of vocal students requires a more subjective assessment (Helwig & Thomas, 1973). Researchers have identified programmatic and teacher factors that facilitate student achievement within successful 11 choral programs, but the researchers did not imply that these factors related to individual student achievement (Levi, 1986; Mudrick, 1997; Wright, 1996). Purpose of the Study The primary purpose of this study was to identify the contributions of parental involvement, motivating factors (attributions of success), and SES to performance achievement among high school music ensemble members. The secondary purpose of this study was to compare parental involvement, motivating factors, and SES of high school choir and high school band members. The differences in performance demands and ensemble procedures of instrumental students and choral students might be related to variances in family environment or student attributions. Research Questions The study addresses the following research questions: 1. What are the parental support factors, motivational factors, and SES of high school band and choir students who attend high schools that have both choir and band students selected for all-state participation? 2. How do the factors relate to membership in band or choir ensembles? 3. How do the factors relate to all-state participation in band and choir students? Assumptions Students who have participated in all-state band or chorus in Alabama, Georgia, and Tennessee have auditioned before expert judges. It is assumed that students who have been selected to participate in all-state band or choir have reached a significant level of musical performance ability. Students who have not participated in 12 an all-state festival are not identified as unsuccessful. Many of the students in this category may be successful performers, but for various reasons did not participate in all-state festivals, therefore, they were not identified as successful. Delimitations The researcher chose the sample for this study from a list of schools that have sent students to all-state band and chorus festivals in either 2007, 2008, or both years. The sample, therefore, includes only students who have participated in established, reasonably successful band and choral programs. The factors identified in this study were drawn from high schools through the southeastern United States, but the socioeconomic makeup of the schools in the study is likely to be higher than average. Schools in the least affluent areas of the southeast are not likely as affluent schools to support programs that send a significant number of students to all-state festivals. Parental involvement, personal attributions of success, and SES might differ in schools that have minimal or no band or choral programs. Definitions Selection for all-state band or chorus is a measure of music performance achievement. All-state participation is not a complete explanation of success in band and chorus students, but it reflects an assessment of performance achievement by expert judges. Success in this study is limited to successful performance achievement as indicated by selection for all-state band or choir. 13 A characteristic is a parental involvement variable, an attribution variable, or socioeconomic variable that, in conjunction with other variables, contributes to a factor. Items on the survey identify characteristics or demographic variables. A motivating factor is a group of attribution variables that relate strongly to one another. Asmus identified Effort, Background, Classroom Environment, Musical Ability, and Affect for Music as motivating factors in the Asmus Motivating Factors (AMF) scale (Asmus, 1985a). 14 CHAPTER 2 LITERATURE REVIEW Research on student achievement in music has yielded varying results, because researchers have used varying definitions of success. The methods used to assess music achievement have not always been consistent with the researcher?s definition of music achievement. Interest in the topic of the current study began with readings about characteristics of expert performers. Woody (2001) sought to apply the findings of research on expert performers to music education. Woody identified practice habits and motivation that were associated with advanced performance abilities and he pointed out that parental support was related to the development of these habits. The current researcher?s area of interest is vocal music education. The initial concept of the study was to examine successful choral students in the same manner that Woody examined expert performers. The literature review for the current study began with a search of articles that included the terms vocal success, vocal, chorus, choral students, successful music students, and music student characteristics. The researcher began the search in journals related to music education research and the Handbook of Research in Music Education (Colwell, 1992). The search yielded few studies that focused on choral students. Most of the studies on successful choral students or characteristics of choral students were based on successful choral programs 15 or characteristics of choral teachers or choir directors. Most articles on vocal performance and vocal success were reviews of performances and performers. The search revealed a body of research on successful instrumental and general music students related to student attributions of success. The studies based on attribution theory and student motivation pointed to the importance of the environment of the student. The researcher then concentrated on studies of parental involvement and home environment on student success. The terms factors and predictors of success were used more commonly in the studies than characteristics of success. A search for successful instrumental students produced a compilation of studies on factors related to performance achievement in junior high and high school students. Similar studies on vocal performance achievement were rare. The researcher examined the studies of instrumental students to identify methods that would appropriately identify factors related to performance achievement in choral students. A common concern among researchers was the relationship of SES to musical achievement. The researcher investigated socioeconomic measurements used in other studies and used the Handbook of Research Design and Social Measurement (Miller, 1991) to assess the value of these measurements for the current study. Overview of Research into Predictors of Musical Achievement The literature reviewed for this dissertation included studies of an array of variables that have been identified as predictors of achievement in music. Four prominent categories of predictors emerged in the review of the literature. The first category included IQ and musical aptitude (Helwig & Thomas, 1973; Hufstader, 1974; 16 Klinedinst, 1991; Kuhlman, 2005; Manor, 1950; McCarthy, 1974; Young, 1971). The second category, described as home environment, includes the physical attributes of the home (Brand, 1986), parental involvement in the musical experiences and musical education of the child, and musical background of parents and siblings (Bonifati, 1997; Brand, 1985, 1986; Br?ndstr?m, 2000; Davidson & Borthwick, 2002; Davidson, Howe, Moore, & Sloboda, 1996; Sloboda & Howe, 1991; Zdzinski, 1992, 1993, 1996). The third category of predictors was SES (Albert, 2006; Br?ndstr?m, 2000; Dunlap, 1975; Fitzpatrick, 2006; Klinedinst, 1991; McCarthy, 1980) and the fourth category of predictors was motivation and attribution theory (Asmus, 1985a, 1985b, 1986a, 1986b, 1989; Asmus & Harrison, 1994; Legette, 1993, 1998, 2003; Schmidt, 2005). The predictors of music achievement have been most frequently studied in relationship to success in instrumental students. Music achievement, intelligence, and musical aptitude Kuhlman?s (2005) overview of the research into predictors of musical achievement included studies on the relationship of IQ and musical aptitude to student achievement in music. Many researchers have identified musical aptitude and academic ability as the two strongest predictors of success in instrumental music. Manor (1950) suggested that IQ measurements (IQ) be discarded from the battery of tests used to predict instrumental success, but few researchers since have disregarded IQ as a viable predictor of instrumental success. Klinedinst (1991) identified scholastic achievement and academic achievement as the strongest predictors of performance success in fifth-graders? first year of 17 instrumental study. Klinedinst?s study of the possible predictors of students? performance success included musical aptitude, scholastic ability, academic achievement, attitudes toward music, self-concept in music, music background, achievement motivation, SES, and physical characteristics of the students. Klinedinst used stepwise multiple regression, discriminant function analysis (DFA), and Pearson product-moment correlation to identify relationships among the predictors. Success was measured by teacher rating, performance rating by a judge, and student retention. Scholastic ability in math and reading achievement were the strongest predictors of musical success (24%), as measured by the teacher ratings of musical ability. Although music aptitude was found to be a statistically significant predictor of success, it accounted for less than 10% of the variance in music achievement. The results of Klinedinst?s study suggest that music teachers should find school academic records and music aptitude testing to be valuable in their recruitment for music students. The testing and evaluation procedures used in Klinedinst?s study may be valuable for diagnosing current students. Knowledge of one?s students facilitates the adaptation of instruction for individual students or for the whole class (Klinedinst, 1991). Hufstader (1974) found that musical aptitude, musicality, and musical intelligence variables provided an 85% prediction rate of fourth through sixth-grade students who would be successful in instrumental music. Hufstader used the California Test of Mental Maturity (CTMM) and the California Achievement Test (CAT) to collect data on IQ and musical aptitude. The band directors of the study group 18 identified successful and unsuccessful students by ranking them in their class based on the students? technique, tone quality, musical reading ability, rhythmic reading ability, and general musicianship. Data from the highest 33% of the students and the lowest 33% of the students were used in the statistical analysis. A DFA of each item in the CTMM and the CAT revealed that every variable in the tests provided a unique contribution toward identifying successful and unsuccessful students. One concern over the validity of the results in Hufstader?s study was the small number of participants (n = 34). Each of the four classes was ranked by a different teacher, but the criteria used to rank the students were specific and objective. No comparison was made between classes; therefore, this ranking may not completely reflect the differences in abilities between classes. Based on the score profiles of the two groups, four participants were deleted from the high group and one subject was deleted from the low. Young (1971) examined the functions of IQ, academic achievement, and musical aptitude as factors in predicting instrumental music achievement in fifth-grade students (N = 709). Young measured IQ with the Lorge-Thorndike Intelligence Test, academic achievement with the Iowa Tests of Basic Skills (ITBS)), and musical aptitude with the Musical Aptitude Profile (MAP). Students participated in small group (2-5 students) lessons for seven months. Musical success was determined through evaluation of a recorded performance at the end of the seven months. Young found the highest correlation between musical success and the predictor tests when all three predictor tests were combined. Young found that the measures of success were 19 predicted as successfully with only the MAP and ITBS scores as they were with all three tests. The ITBS scores alone were only slightly less accurate in predicting musical success than the MAP and ITBS scores combined. The students who prematurely dropped out of the instrumental program scored lower overall in all the tests than the students who remained in the program. However, the IQ scores of those who started and dropped out were still higher than the scores of the general student population. A greater difference existed in the ITBS scores than in the IQ scores between the group that dropped out and the group that completed the seven months. Again, Young found that the group that dropped out scored higher on each test than the general student population of the same grade level implying that regardless of the role of IQ as a factor in musical success, it is not a predictor of longevity in the instrumental program. The MAP scores of those who dropped out were lower than the general student population. The rhythm aspects of the MAP tests appeared to be the most significant factors related to dropping out of the program early. Young determined that each test successfully predicted success in specific facets of musical achievement. The best predictor of success in all areas of musical achievement was the composite score of all three tests. Young?s findings illustrate the difficulty in identifying success in music and choosing factors to measure success. High student scores in musical aptitude criteria tests correlated positively with musical abilities unrelated to reading. Young found that high student academic ability and IQ correlated positively with music skills related to music reading and notation. Facets of musical achievement that did not 20 require music reading skills revealed a strong correlation with the MAP. Facets of achievement that required music-reading skills did not reveal a strong correlation with the MAP scores. Young noted a weak correlation (r = .23) between the sight-reading and improvisation skills of the students. The finding was similar to the correlation found between IQ and musical aptitude (r = .25). Young concluded that improvisation and music reading skills are unrelated skills. McCarthy (1974) reported a high correlation of IQ and academic success among seventh-grade beginning instrumental students (N = 90). The primary purpose of McCarthy?s study was to create and evaluate a tutorial instructional method that would facilitate student learning in instrumental music. The method would account for the differences in the physiological and psychological development of the students. McCarthy tested the individual tutorial instruction by using a control group (n = 45) and an experimental group (n = 45). The students in the experimental group were given individual instruction and assignments within class and were individually evaluated. The control group was taught with a traditional ensemble approach. In the control group, performance achievement correlated positively with attitude towards musical and personal adjustment. In the experimental group, these same variables exhibited almost random relationships. Performance achievement was higher in the experimental group, indicating that the individual instructional approach was beneficial to student success. Investigating the effect of IQ and grade point average (GPA) on performance achievement was a secondary purpose in the study. The results 21 indicated that students in both groups who had the highest IQs and GPAs scored highest in performance achievement. McCarthy (1980) evaluated the relationship of individual instruction to achievement and dropout in fifth- and sixth-grade students (N = 1199). By measuring music achievement, SES, and IQ, McCarthy found that student?s academic reading level accounted for over 85% of the variance in their sight-reading ability as measured by the Watkins-Farnum test. Helwig and Thomas (1973) evaluated studies that identified predictors of musical success. They noted that little evaluation of student potential and progress existed in choral music studies, and, recommended that choral teachers should evaluate students according to ability and place them in the appropriate level of chorus in order for the students to be successful. Placement above or below the student?s ability level may detrimentally affect the student?s success. Helwig and Thomas?s purpose was to determine if musicality and IQ scores would more accurately predict a student?s success in a vocal performance class than traditional auditions and observation methods. The secondary purpose was to identify teacher bias in the grading. Helwig and Thomas found support for the use of IQ and musical aptitude scores to predict students? achievement in 10th- through 12th-grade choral classes. The researchers found that success in choral achievement could be predicted using the Gaston Test of Musicality and the CTMM. The researchers operationalized students? success through teachers? grades based on attitude, effort, and musicality. Helwig and 22 Thomas described the correlation between the IQ and musicality measurements and the grades of the students, but the discussion of the secondary purpose overshadowed their description of results. They did not explain why only 64 participants were selected out of a possible 286 participants. IQ and musical aptitude appear to be excellent predictors of musical ability if the appropriate measurements are properly matched to the type of musical ability that is being measured. If the role of an educator is to predict success, then IQ and aptitude are of value. However, IQ and musical aptitude are internal, stable attributes and are of minimal benefit toward improving students? musical ability. The students? environment is an external, unstable attribute that can be shaped to improve students? musical ability. Family Background Studies on family background include information about parental support in music and family characteristics. Family characteristics encompass the parents? musical backgrounds, siblings, and demographics such as SES and geographical location. Just as the home environment appears to have an effect on a variety of human characteristics such as school achievement, IQ, student attitudes and expectations, and creativity, it might have a significant effect upon musical development (Brand, 1985). The Home Musical Environmental Scale (HOMES) developed by Brand (1985) was designed to evaluate the characteristics of homes that provide positive environments for musical development. Brand included 15 items in the HOMES 23 questionnaire. The items examined parental musical backgrounds, parental participation with their child?s musical activities, parental concert attendance, and parental provisions for the home musical environment (providing musical supplies, instruments, and listening devices). Brand intentionally chose a homogenous socioeconomic sample for the study so that the survey would provide data concerning factors that are changeable, as opposed to SES, parent education, and parent occupation. Most participants were Hispanic students of low or low-middle SES. Brand identified four factors that accounted for 63% of the variance of home environment in second grade students? achievement. In a factor analysis the variables of parents? attitudes toward music and musical involvement with their children, parents? concert attendance, children?s ownership and use of records and tapes, and parents? ability to play a musical instrument were all identified as significant factors (Brand, 1985). In a subsequent study, Brand (1986) used HOMES to investigate the correlation between home musical environment and musical attributes of 116 children age 7. Brand used a sequence of multiple regression analyses to estimate the relationships of each of the environmental factors to each of the variables in the Primary Measures of Musical Audiation (PMMA). The composite of all the environmental factors accounted for 20% of the variance in PMMA scores. Brand emphasized the significance of the home environment within a homogeneous socioeconomic setting. The study points out that the home environment, even within a 24 low socioeconomic setting, has a significant relationship to the student?s musical achievement. Bonifati (1997) investigated the impact of home environment on the success of instrumental students in grades 4 - 12. Most instrumental students who were identified as successful by their teacher took lessons on their instrument and owned their instrument rather than renting it. Most of the students came from two-parent households. Parents were between the ages of 36 and 50, had college degrees, professional occupations, were typically white, and were Protestant or Catholic. Parents? musical experience had little relationship to the students? success, but parents? support for their child?s musical endeavors was positively related to their child?s success. Bonifati identified parental encouragement as the most important factor in their child?s success. Most parents did not want a music career for their child, but they expected their child to be committed to continuing music studies. Most of the students began music studies when they were less than 5 years old. Many had taken piano lessons, were generally successful in their academic endeavors, and had other siblings who were successful in music. Davidson, Howe, Moore, and Sloboda (1996) interviewed children (N = 257) ages 8 through 18 who had received instruction in instrumental music. The researchers divided the students into five groups for the study. Group 1 (n = 119) included students who attended a music school and anticipated making a career in music. Group 2 (n = 30) included students who were called for an audition, but were not admitted to the music school. Group 3 (n = 23) included students who inquired about the music 25 school, but did not apply. Group 4 (n = 27) included students who attended public school and learned to play an instrument, but did not intend to make music a career. Group 5 (n = 58) included students who attended the same public school but had discontinued playing an instrument at least a year prior to their interview. Davidson, et al. interviewed all of the students and at least one of the parents of each student. The researchers found a trend in the parental support over a period of years. The strongest parental support for the students in group 1, took place before the students were 11 years old. Parents of group 1 encouraged singing before age 3. As the students? age increased, their self-motivation and autonomy increased and their parental involvement decreased. The students in groups 4 and 5 did not receive early parental support. In groups 4 and 5, parental pressure for students to practice increased during the students? teenage years. None of the groups indicated a particularly noticeable musical interest at an early age. Parents of group 1 were the most involved in music, albeit at an amateur level, and parents of group 5 were the least involved in music. Davidson, et al., asserted that parental involvement is critical to student success in music and that this involvement must begin in the preschool years. In a longitudinal case study, Davidson and Borthwick (2002) followed the family dynamics of an English family for 13 months. One researcher was integrated into family activities in order to explore family dynamics in detail. The other researcher visited the family every two weeks to observe the family with a more objective perspective. Davidson and Borthwick (2002) found that children who eventually became professional instrumental performers were not only monitored by 26 their parents, but had varying complex interactions with their parents. Parental support was found to be important to the motivation and self-worth of a child. The amount and the nature of parental involvement in a child?s musical development were also important. Davidson and Borthwick concluded that the type of support provided by parents can shape a child?s success or failure in music and that parent?s expectation of a child will be reflected in the child?s own self-expectation. Differences in levels of expectation between children may result in differences in self-worth; a child with lower levels of expectations will often have lower self-worth. Parents must find a balance between responsiveness (warmth and acceptance) and demandingness (controlling and restrictive) (Davidson & Borthwick, 2002). Br?ndstr?m (2000) observed that musical background and SES appeared to be significant factors in children?s musical activities. Br?ndstr?m investigated 12- and 13-year-old students in sixth grade (N = 369) at 11 different schools in Sweden. Six years later, Br?ndstr?m sent a questionnaire to one of the classes from a participating school to determine how long the participants (N = 13), now in their last year of school, had studied in the Municipal Music School, what instrument they had studied and what occupational choice they had made. Br?ndstr?m measured the effect of SES, musical background of the family, students? plans for their future, and students? choice of instrument. These variables were compared to three groups: those studying music currently (n = 90), those who had studied but discontinued music (n = 147), and those who had never studied music (n =132). Fifty-four percent of the children who continued their studies in the Municipal Music Schools in Sweden had parents who 27 currently or previously played a musical instrument. Fifty-eight of these children had siblings who played a musical instrument. Br?ndstr?m reasoned that a parent?s interest in music facilitates the tangible help that the parent can give their child at home. Pitts, Davidson, and McPherson (2000) examined motivational, personality, and environmental characteristics of nine primary-school students in their first 20 months of instrumental studies. The method of data collection was a collection of longitudinal case studies selected from 158 brass and woodwind players during the 20- month period. The nine participants attended eight different primary schools. The study consisted of three groupings of students. Three of the students (group A) had maintained interest and enthusiasm for their instrument, three of the students (group B) continued to take lessons beyond the 20 months, but with decreased motivation, and three of the students (group C) discontinued music lessons within the 20 months of the study. Motivation in group A tended to be intrinsic even though it included extrinsic factors. The children set high standards for themselves. The children in groups B and C were motivated only by extrinsic factors. In groups B and C, practice time was a factor more than practice quality. In some cases respondents said they ?put in the time? and in some cases, there was no practice focus at all. Parents of group A encouraged their children and helped them to set realistic expectations. Parent of groups B and C often exhibited limited involvement and either limited or unrealistic expectations of their children. They were often distanced from the child?s playing and insincere or injudicious in their praise to the child. Even motivated children 28 experienced periods of self-doubt and low interest. Parental and teacher support was critical to overcoming those periods (Pitts, Davidson, & McPherson, 2000). Zdzinski (1992, 1993, 1996, 2007) investigated the relationship of aspects of parental involvement with music aptitude, musical achievement, and performance achievement in instrumental studies. Zdzinski (1992) used a researcher developed Parental Involvement Measure (PIM) instrument and HOMES to identify parental involvement in middle school instrumental students (N = 113). Zdzinski patterned the PIM after Doan?s (1973) Measurement of Family Involvement in Music (FIM) and HOMES. PIM included three sub-scales. The Parental Involvement-Frequency (PI-F) sub-scale consisted of 15 five-point Likert-type questions that measure the frequency with which parents are involved in the musical activities of their children. The low range of the PI-F (15 points) indicated no parental involvement in the child?s musical activity. The high range of the PI-F (75 points) indicated the highest measurable frequency of parental involvement. The Parental Involvement-Degree (PI-D) sub-scale consisted of 15 questions that measure the degree of involvement by the parents (father only, mother only, or both). The low range of the PI-D (0 points) indicated no involvement of either parent. The high range of the PI-D (30 points) indicated the involvement of both parents in all areas of musical activity included in the survey. The Parental Involvement-Categorical (PI-D) sub-scale consisted of nine parental involvement items with yes/no responses that identified home environment characteristics. 29 To establish content validity in the PIM items, Zdzinski (1993) solicited responses from high achieving wind instrumentalists and instrumental music teachers about parental involvement that related to high student achievement. Cronbach?s index of internal consistency was r = .94 and test-retest reliability was r = .85. Zdzinski added nine yes/no demographic questions about gender, grade, age, school, years of playing experience, practice time, and private instruction. Performance achievement correlated with six PIM variables, indicating that parents take the students to concerts (r = .251), attend non-school concerts (r = .227), provide transportation (r = .192), play in a musical group (r = .171), attend rehearsals (r = .165), and listen to music (r = .155). Music achievement correlated with six PIM variables: parents talk about music (r = .321), listen to music (r = .261), take the students to concerts (r = .260), play in a musical group (r = .199), provide transportation for the student (r = .188), and attend parent meetings (r = .158). While the correlations between parental involvement and music achievement were statistically significant, they were too low to suggest practical value (Zdzinski, 1992). Zdzinski (1996) expanded the investigation on the relationships between parental involvement, music aptitude, grade level, and gender to performance, cognitive achievement, and the progression of musical attitudes in instrumental students (N = 406) in grades 4 through 6. The results supported Zdzinski?s (1992, 1993) earlier findings concerning parental support. Parental involvement significantly correlated to affective, cognitive, and performance results. The parental involvement correlation to affective measurements increased as the grade level increased. The 30 correlation between parental involvement and affective measurements accounted for 12.9% of the shared variance. Parental involvement correlated to cognitive and performance measurements in the elementary grades, but not in the middle school and high school grades. Parental involvement correlated to the musical achievement of instrumental students. The relationships were, however, too small to justify practical value (Zdzinski, 1996). Zdzinski (2007) studied 523 elementary, middle, and high school students in general music, orchestra, band, and chorus classes to determine the reliability of the 39-item Parental Involvement-Home Environment in Music (PI-HEM) scale. Using a principal components analysis with a varimax rotation, Zdzinski identified home structure, parental expectations, musical participation, musical environment, family musical background, and attitudes about music as factors related to all the groups. The ratio of participants (N = 523) to variables (36) was more than 14.75 to 1. The result of the Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was .85. When Zdzinski administered the 99-item PI-HEM, the ratio of participants to variables was more than 5:1and the KMO measure of sampling adequacy was .88. The analysis of the 99-item scale revealed a seventh factor; parental program support. Sloboda and Howe (1991) sought to understand factors that related to high levels of competence in instrumental performance. Their study was based on references to the relevance of family background in the success of Nobel Prize winners, champion chess players, and prize-winning scientists. The authors interviewed 42 students, age 8 - 18, from Chethams School in Manchester, England 31 and 20 of the students? parents. The students were identified by the school as A-level or B-level musicians. Sloboda and Howe found a relationship between parental support and student success. Student success did not appear to relate to personal involvement of the parent in musical activities. Most of the parents were involved in their child?s progress. Most of the children required a significant amount of encouragement from their parents to practice. Six of the students began instrumental studies based on their internal, personal motivations. Five of these were identified as having extraordinary abilities on their instrument (Sloboda & Howe, 1991). Sloboda and Howe suggested that parents provided a balance between placing extreme pressure on their child to practice and letting them practice based on their internal motivation. Socioeconomic Status Albert (2006) noted a relationship between SES and student?s motivation for success in school. Albert suggests the same relationship is possible between SES and students? participation in instrumental music programs. The cost of participation in instrumental programs may be a deterrent to the initial participation and the continued participation of students from low socioeconomic backgrounds (Albert, 2006). Frakes, (1984) found that the dropout rate in choral programs was higher than the rate in band programs and that most of this took place in junior high. Tipps (2003) suggested that once the financial investment in an instrument was made students were more likely to continue in the program. Bonifati (1997) interviewed the parents of 19 instrumental students (age 9 - 13) who were identified by the students? music teachers as successful. Bonifati identified 32 the parents? musical background, musical activities, and SES through questionnaires and interviews. Bonifati did not find a correlation between SES and success in instrumental studies, but this may have been due to the small number of participants in the study. Conversely, Klinedinst (1991) found SES to be a significant factor in predicting instrumental performance achievement and retention in instrumental programs for fifth-grade students. Klinedinst used the Hollingshead two-factor index to identify SES and used teacher rating (based on a scale developed by Klinedinst) to measure students? potential for success. Even though scholastic ability in math and reading achievement was the strongest predictor of musical success, SES was the strongest predictor of retention (F = 6.82), followed by self-concept in music (F = 5.41), reading (F = 4.39), scholastic ability (F = 3.56), and math achievement (F = 3.17). Br?ndstr?m (2000) found that 12- and 13-year-old children of parents with middle and high-level professional occupations and high-academic backgrounds were approximately twice likely as children of manual laborers to attend a Municipal Music School, thereby supporting the conclusion that SES has an important relationship to musical success. Fitzpatrick (2006) measured the effects of SES and instrumental performance in students (N = 15,431) grades 9 - 12 in the Columbus Ohio Public Schools. Student SES was identified by free and reduced lunch records. The researcher obtained the Ohio Proficiency Test (OPT) scores in writing, reading, mathematics, science and citizenship for all students in grades 4, 6, and 9. Fitzpatrick divided the students into two groups according to instrumental and non-instrumental participation. The OPT 33 scores from grade 4 were linked to the students? future instrumental or non- instrumental status. The researcher compared the students SES to their OPT scores. In 7 of the 12 OPT sub-tests, high SES students outscored all other students. Instrumental students outperformed non-instrumental students in all areas for students of similar SES. Instrumental students of similar SES began with higher OPT scores. By ninth grade, instrumentalists of low SES outscored non-instrumentalists of high SES. Dunlap (1975) explored the effect of SES, race, community size, and the presence of a father in the home on the musical achievement of students (N = 472) in Mississippi and Arkansas. Dunlap found that SES correlated positively with music achievement among all the students of the sample and among students in each of the sub-groups. The sub-groups in Dunlap?s study were black students, urban students, and metropolitan students. Dunlap based his measurement of SES on Warner?s (1960) four-aspect index. Warner identified SES through occupation, income, housing, and dwelling area. Each of the four aspects was measured on a 7-level scale. Warner then weighted each aspect to determine a family?s socioeconomic index. Warner weighed occupation by 4, source of income by 3, house type by 3, and dwelling area by 2. Socioeconomic measurement instruments Sociologists have grappled with the problem of vague identifications of occupations versus overly specific identifications of occupations (Van Leeuwen, Mass, & Miles 2004). Goyder and Frank (2007) created codes for occupational status based on skills involved in the occupation. The nine skill-types developed by Goyder and Frank were 1) business, finance and administrative occupations; 2) natural and 34 applied sciences and related occupations; 3) health occupations; 4) occupations in social science, education, government service, and religion; 5) occupations in art, culture, recreation, and sport; 6) sales and service occupations; 7) trades, transport and equipment operators, and related occupations; 8) occupations unique to primary industry; and 9) occupations unique to processing, manufacturing and utilities. Two problems existed in using this scale. First, the scale was too vague to place all of the data accurately and confidently. Second, the scale utilized Canadian data that did not follow the criteria generally used in studies in the United States. Osborn (1987) observed that before 1950 SES was commonly measured by the occupational status of the male head of the household. The Hollingshead Index of Social Position and the Duncan Socioeconomic Index have been widely used to identify SES (Osborn, 1987). The most notable problem in Duncan?s index is poor reliability when occupational descriptions are improperly converted into occupational codes. Researchers need to be significantly trained to implement the complex coding system of the U. S. Census Index of Occupations and Industries and the Dictionary of Occupation Titles from the U. S. Department of Labor (Miller & Salkind, 2002). Deonandan, Campbell, Ostbye, Tummon, and Robertson (2000) compared seven socioeconomic measures that were based on occupation, education and income. Four of the measures (Blishen, Pineo-Porter, Hollingshead, and British) used self- reported data. Three measures relied on estimates of occupation, income, and education based on zip code information (Deonandan, et al., 2000). Deonandan?s, et al., concern with the self-reporting measures was the investigators? subjectivity in 35 categorizing the occupations. Deonandan, et al., found a high correlation between the four self-reporting measures, but a low correlation between the self-reporting and the postal-code measures. Researchers have reported high reliability and validity in Hollingshead and Redlich?s (1958) measure of SES. The Hollingshead three-factor index consists of occupation, education, and residence scales. In order to use the residential scale, the researcher must map the geographical area that encompasses the research participants into residential zones. The subjectivity of rating neighborhoods and the impracticality of mapping residential zones caused the three-factor index to be less widely received by researchers than the two-factor index (Osborn, 1987). Hollingshead?s two-factor index includes an occupational scale and an educational scale (Miller & Salkind, 2002). Hollingshead?s two-factor and three-factor indexes correlate highly with judged class status. The two-factor index correlation is r= .9406 and the three-factor correlation is R = .942 (Hollingshead & Redlich, 1958). Hollingshead and Redlich?s measures of social position show a strong correlation to Ellis, Lane, and Olesen?s (1963) index of class position (Miller & Salkind, 2002). Longitudinal studies that have used the Hollingshead occupational status index indicated that Hollingshead?s scale is as strong as any of the other prominent scales of SES (Slomczynski, Miller, & Kohn, 1981). The Hollingshead scale was reprinted in the Handbook of Research Design and Social Measurement (Miller & Salkind, 2002, p. 462-469) with updated financial information on the occupational scale. Hollingshead categorized nearly 200 36 occupations into seven levels of occupational status. Hollingshead identified seven levels of educational status in the education scale (Miller & Salkind, 2002). The highest occupational level in the household and the highest educational level in the household represented the occupational and educational levels of each household (Davis-Kean, 2005). Table 1 describes the seven levels of the occupational and educational scales. In Hollingshead?s two-factor index, occupation is weighted by seven and education is weighted by four. Hollingshead categorized the total of the weighted scores, based on the population of New Haven, CT into five social classes. Class I was the lowest socioeconomic class with a range of 11-17. The range of class II was 18-31, class III was 32-47, class IV was 48-63, and class V was 64-77 (Miller, 1991). 37 Table 1 Hollingshead?s Occupational and Educational Scale Level Occupation Education 1 Higher executives of large concerns, proprietors, and major professionals Graduate professional training 2 Business managers, proprietors of medium- sized businesses, and lesser professionals Standard college or university graduation 3 Administrative personnel, owners of small businesses, and minor professionals Partial college training (completed at least one year) 4 Clerical and sales workers, technicians, and owners of small businesses High school graduation 5 Skilled manual employees Partial high school (at least 10th grade) 6 Machine operators and semiskilled employees Junior high school (7th-9th grade) 7 Unskilled employees Less than 7 years of school Motivation and Attribution Theory According to Asmus (1985a) the effect of motivation upon musical achievement is poorly understood. An inequality of motivation exists in many classrooms. In order to promote student motivation, teachers must understand what 38 motivates their students. Students? appearance of laziness, weak character, and shortsightedness are commonly misdiagnosed as symptoms of apathy. Teachers need to be able to differentiate between causal attributions of success or failure and perceived characteristics of apathy (Legette, 2003). Greenberg (1970) identified low self-concept as the primary cause of low musical achievement. Greenberg?s study indicated that low achievement in pitch matching was not a result of a musical factor, but an emotional or psychological factor. Hylton (1981) conversely suggests that musical experience appears to create a positive self-image. Attribution theory suggests that students consider ability (internal-stable attributions) and effort (internal-unstable attributions) to be the primary reasons for success and failure in music among elementary, middle, and high school students (Arnold, 1997; Asmus, 1986a, 1986b; Legette, 1998, 2003). In a 4-year longitudinal study, Rathunde and Csikszentmihalyi (1993) assessed high school freshmen?s (N = 208) performance in math, science, music, and art. Undivided interest appeared to be a factor for success in academics (math and science) and performance (music and art). Undivided interest is a concept that Dewey (1933, p. 209-213) described as being ?playful and serious at the same time.? Students must enjoy a task while having a goal. Piaget (1962, p. 168) observed that play satisfies the ego and provides sensory-motor or intellectual satisfaction. Rathunde & Csikszentmihalyi (1993) asserted that assimilation of a task is accomplished through repetition. Piaget (1962, p. 182-192) described assimilation as the incorporation of new information into one?s current knowledge. Rathunde & Csikszentmihalyi referred 39 to the assimilation process in terms of mastering a task. Interest must be present in order for this repetition to take place. Rathunde and Csikszentmihalyi adjusted for the effects of family support and income, scholastic aptitude, achievement orientation, and gender in their data. They found a positive correlation between undivided interest and talent mastery, undivided interest and teacher ratings, and undivided interest and subjective engagement. Students with multiple talent areas may regard each talent area differently. They may focus on one talent area to the exclusion of another simply because of time limitations. To correct for this, the researchers deleted all data from multiply-talented students and performed the same statistical analysis. The researchers found that the resulting data were almost identical to the data that was analyzed with the inclusion of multiply-talented students. Asmus (1985a, 1985b) utilized the concepts of attribution theory to understand elements of students? success or failure in music. Asmus presented a two-question survey to 118 sixth- grade students. The participants attended three different schools. Twenty of the participants attended a middle class parochial school, 55 of them attended an inner city school, and 43 of them attended an affluent suburban public school. The students identified five reasons they believed some students do well in music. In the second question, the students identified five reasons they believe some students do not do well in music. The responses were categorized according to one of four causal categories: ability, task difficulty, luck, or effort. The attributions associated with the categories were stable or controllable (ability and task difficulty), unstable or uncontrollable (luck and effort), internal (ability and effort), and external 40 (task difficulty). The results did not indicate a significant difference in the attributions for success and the attributions for failure. Using a 3 x 2 repeated measures multivariate ANOVA, Asmus found a significant difference in attributions between the schools. Students at the parochial school identified ability as the primary cause for success or failure. Students at the inner city school identified effort as the primary cause. Students at the suburban schools identified ability and effort as the primary causes. Students at all schools identified luck as a reason for failure, but the parochial school students identified luck less than students in the other two schools. Asmus was surprised that internal-stable attributes were not more prevalent, because society often considers musical skill to be a ?gift.? Asmus (1986b) expanded the study of attribution to include students (N = 589) in grades 4 - 12 in instrumental, vocal, and general music. Females ascribed internal- stable attributes to success and failure more than males did. Students tended to ascribe stable attributions to success and external-unstable attributes to failure. As grade levels increased, students shifted from internal-unstable attributes to internal-stable attributes. The shift from effort related to ability related success indicated a decrease in persistence in the older students. The study implies that that teachers need to encourage students with effort-related attributions (Asmus, 1986b). The Asmus Motivating Factors Measure (AMF) is one of two measures that comprise the Asmus Measures of Motivation in Music (AMMM) (Asmus, 1985a). The AMF measures Effort, Background, Classroom Environment, Musical Ability, and Affect for Music. Asmus developed the AMMM by collecting more than 5,000 41 statements from music students in 4th through 12th grades. A different group of high school music students (n = 540) rated the appropriateness of the 125 most common statements about success and failure in music (Asmus, 1986b). Through a factor analysis, Asmus identified the five factors, related to success in instrumental students, which comprised the AMF. The factors were Effort, Background, Classroom Environment, Musical Ability, and Affect for Music. Reliability for the five scales ranged from .60 to .90 (Asmus, 1986a, 1987, 1988, Asmus & Harrison, 1994). Asmus used the same process to develop three scales comprising the Motivation Magnitude Measure, which is the second measure in of the AMMM (Asmus, 1989). Asmus compared a teacher ranking of the students with the results of the AMMM to establish criterion-related validity. Criterion-related validity was low. Asmus questioned whether using a teacher ranking was a suitable criteria to establish validity as teachers might use significantly different criteria to measure motivation than the students do, as indicated on their responses on the AMMM. To establish construct validity Asmus compared the factor analyses of the individual motivating factors scales and magnitude of motivating scales to the AMMM. Construct validity of the AMF was strong in that the factors of the motivating factors scales were identical to the factors on the AMMM. The factors found in the magnitude of motivation scale revealed low construct validity in relation to the AMMM. Reliability for Asmus? AMF measure was ? =.728. Asmus found that the stability dimension was more difficult to define in the external and emotional dimension than the internal dimension of affect 42 for music. This might be because the classroom environment that is related to external dimensions fluctuates more than the affect for music Legette (1993) examined the effect of Effort, Background, Classroom Environment, Musical Ability, and Affect for Music on college students? (N = 105) success in a beginning guitar class. Legette used the 35 item Music Attribution Orientation Scale (MAOS) (Asmus, 1988) to investigate the differences between music majors (n = 43) and non-music majors (n = 62). The MAOS contained five sub- scales identified as Effort, Background, Classroom Environment, Musical Ability, and Affect for Music. Both groups collectively and each group individually placed the most importance on Effort, followed in order, by Affect for Music, and Musical Ability. Music majors placed more importance on each of the three attributions than non-music majors did. No statistical difference existed between music majors and non-music majors for the causal attributions of class environment and background. The study confirms Asmus? (1985a, 1985b, 1986b) findings that students attribute ability and effort to success and failure in music. The researcher unexpectedly found that the non-music majors often performed better than the music majors did. Legette speculated that music majors did not believe that intense effort and ability were necessary in this class. They were not as concerned about performance and skill acquisition in this particular class. Non-music majors may have viewed the class as an opportunity to develop a new skill. Legette (1998) found that high school students (N = 1,114) identified musical ability as the strongest attribute (M = 4.12 on a 5 point Likert-type scale) for success. 43 Using the Asmus (1988) MAOS, Legette investigated Effort, Background, Classroom Environment, Musical Ability, and Affect for Music. The second strongest attribute was Effort (M = 4.04). Legette analyzed the differences due to school system and gender through a t test for two independent samples. Students in city schools indicated significantly higher attributions of success for Effort, and Musical Ability (p < .02) than students in county schools, but significantly lower for Class Environment (p < .02). Attributions of Effort, Background, Musical Ability and Affect for Music, increased significantly from elementary to middle school and from middle to high school (p < .02). Females indicated higher attributions for success in Effort, Background, Class Environment, and Affect for Music (p < .02) while males indicated a higher attribution for success in Musical Ability (p < .02). Legette?s results were consistent with earlier findings of students? emphasis on ability and effort. The fact that females emphasized effort more than males conflicted with earlier studies that identified females as more external than males in their attributions. The analysis revealed that students attending city schools placed more importance on Effort and Music Ability, but the students from the county schools placed more importance on Class Environment. One-way ANOVAs revealed no significant differences between school levels for Class Environment, but the students? attributions of Effort, Background, Musical Ability, and Affect for Music increased as the school grade level increased. Legette concluded that teachers should note the importance of ability and 44 effort as causal attributions for success and failure. Students who perceive ability as a cause for failure will be likely to expect continued failure in music despite their effort. Legette (2003) compared the attributions of students in grades 3 through 5 (N = 301) in two contrasting elementary schools. Students in school A were 95% Caucasian, 3% African American, and 18% free or reduced lunch. Students in school B were 15% Caucasian, 75% African American, and 80% free or reduced lunch. Students in both schools indicated that Effort and Musical Ability were the most important causal attributions for success and failure in music. Males indicated more importance for Effort and Affect for Music than females did. School B ranked background and affect for music as more important than school A did. Legette maintained that the perceived importance of effort should be encouraging for teachers. However, Legette warned that if teachers treat effort alone as a cause for failure, then some students who have tried hard, but not been successful, could become even more discouraged. Schmidt (2005) examined the achievement orientation of 300 band students. Schmidt?s variables were task/learning and performance/ego. Schmidt investigated the relationships among achievement orientations, self-concept in instrumental music, and attitude to band compared to teachers? ratings of performance achievement and effort, practice time (self-reported) and demographics, and music experience. Through a confirmatory factor analysis (CFA), Schmidt sought to identify the factor structure that underlay the motivation variables. Schmidt then investigated the relationship of the factors to performance ratings, effort, practice time, music experience, and 45 demographics. Students completed a 5-point Likert-type survey. Schmidt found that most students had a strong musical self-concept. Mastery and cooperative orientations had the highest means and competitive and ego orientations and commitment to band had the lowest means. Commitment to band correlated positively with intrinsic, cooperative, mastery, individual orientations, and self-concept. The competitive and ego orientations correlation, the approach success and avoid failure orientations correlation, and the mastery and intrinsic orientations correlation were all high. Schmidt concluded that proper motivation is critical to student success in instrumental music at all age levels (Schmidt, 2005). All-State Participants Despite the range of measurements for success in music, the opportunities for assessing and predicting student success remains the role of the individual teacher. The inconsistencies in criteria for teacher recommendations of students and the inconsistencies in grading practices for music students cause difficulties in using the criteria for research purposes. The most consistent assessments of students take place at music festivals. State music education associations provide music festivals and competitions as opportunities for student assessment by expert judges. This type of assessment takes place in a setting that provides more objectivity than the classroom setting. Auditions for all-state festivals provide a measurement for individual student performance achievement for choral and band students. Many of the constructs found in the available measurements of musical aptitude and achievement are used in the performance assessment criteria for the all-state festivals. 46 Tobin (2005) chronicled the development of all-state festivals from the 1950s to the present and found little research on the relationship between all-state festivals and the participating students and their music education. Tobin surveyed 727 all-state participants from Massachusetts to investigate the relationship of all-state participation to music, academic, leadership, and extracurricular activities. Tobin found a significant relationship between all-state participation and academic success. Tobin considered the all-state audition process to be rigorous enough to claim that the all- state participants were the best musicians in the state. Lien and Humphries (2001) noted some non-musical factors influenced all- state audition results. All-state bands must accept students based on the number of positions for each instrument in the band. Distance of the students to the audition site seemed to be a factor that resulted in a larger number of students from large cities and large schools auditioning for all-state positions. Ultimately, the selected students from the audition pool were chosen as a result of their success in the audition (Lien & Humphries, 2001). In a study that included 48 states and the District of Columbia, all but one of the states reported holding all-state choral festivals (McCord 2003). Twenty-nine (59.2%) of the state music associations included in the study held live auditions within regions of the state. Eight states (16.3%) auditioned the students at one central location. Four states (8.2%) auditioned the students at a district level and a regional level. Eight states (16.3%) auditioned the students through recorded mediums. Only three of the states in the study used teacher recommendation as the criteria for all-state 47 participation. Most of the states that used judges used one to three judges per audition. Nineteen states (39.6%) provided only one judge per audition, 12 states (24.5%) provided two judges per audition and 16 states (32.7%) provided three or more judges per audition. The highest number of judges used per audition was four: one judge per voice part (SSAATTBB). The number of judges used varied by grade level and the number of voices heard per judge varied when auditioning multiple voice parts in one audition. Sight-singing, foreign language, and scale and arpeggio requirements varied from state to state (McCord, 2003). Summary A review of the research literature concerning factors and characteristics of successful musicians reveals four principal observations. Much of the research of the past 40 years has focused on factors and characteristics of successful instrumental students. A substantial amount of research has investigated musical aptitude in general music students in public schools. A significant void exists in the understanding of factors and characteristics of successful vocal and choral students. Further research is necessary to effectively identify the criteria that define success in music. 48 CHAPTER 3 METHODS AND PROCEDURES Participants The participants in this study were 403 students enrolled in high school choral and band classes in the southeastern United States. The researcher contacted schools in Alabama, Georgia, and Tennessee that had high number of students participating in all-state band and choir. All students in the choral and band classes, with appropriate parental permission, were eligible to participate in the survey. Students who had auditioned and been selected to participate in all-state festivals were identified as successful in performance achievement in music. The sample of all choral and band students in the participating schools represented the population of southeastern high school choral and band students who attended schools with significant participation in all-state festivals. All-State Selection Criteria Audition procedures varied in Alabama, Georgia, and Tennessee, but were similar enough to operationalize successful performance achievement as all-state participation. Alabama, Georgia, and Tennessee maintained state music associations connected to the Music Educators National Convention and state bandmaster?s associations. Alabama and Georgia each maintained a vocal association. The 49 Tennessee Music Education Association (TMEA) maintained three separate vocal associations that were divided geographically into east (ETMEA), middle (MTMEA), and west Tennessee (WTMEA). TMEA provided general guidelines for all-state vocal and all-state band auditions, but the MTMEA and WTMEA had additional guidelines for their own area (Middle Tennessee Vocal Association [MTVA], 2006; Tennessee Music Education Association [TMEA], 2006; West Tennessee Vocal Association [WTVA], (2005). All of the vocal associations auditioned students for performance achievement. Aspects of the audition process and assessment varied among the associations. All of the vocal associations except the Georgia Vocal Association (GVA) required students to learn their vocal part for the selected all-state choral repertoire. All of the vocal associations except the GVA auditioned students in small groups. The number of judges required to audition the small groups or individuals varied among the associations. The AVA, ETVA, and MTVA required that the students audition on their vocal part for the music in the all-state repertoire. The Alabama Vocal Association (AVA) required students to audition without accompaniment for repertoire that is written for a cappella choir. WTVA stated that, ?no student may be required to sing from memory or a cappella? (WTVA, 2005). AVA required only that students prepare their vocal part for the selected choral repertoire. Table 2 illustrates the all-state audition requirements for the vocal associations governing the participants of this study (Alabama Vocal Association [AVA], 2006; Georgia Music Educators Association [GMEA], 2007; MTVA, 2006; TMEA, 2006; WTVA, 2005). 50 Table 2 Vocal Associations? Audition Requirements Vocal Part Judging Sight-reading Scales Tonal Memory Solo AVA 1 judge, knowledge of part, intonation, voice quality Not required Not required Not required Not required GVA Not required Rhythm, intervals, starting and, ending pitch, range Major, natural minor, chromatic Pitch, rhythm diction, tone, interpretation 4 examples 5 notes each ETVA 2 judges per quartet 4 judges per octet Not required Not required Not required Not required MTVA 5 judges selected from participating teachers, blind audition Not required Not required Not required Not required WTVA 3 judges, diction, technical accuracy, pitch, rhythm, Starting pitch, correct notes and rhythms Not required Not required Not required 51 Bandmaster?s associations in Alabama, Georgia, and Tennessee assessed student performance achievement in the all-state festival auditions. The bandmaster?s associations required major and minor scales, arpeggios, sight-reading exercises, and prepared studies, exercises, or etudes of all auditionees. Percussionists were required to audition on snare, xylophone, and timpani (Alabama Music Educators Association [AMEA], 2007; GMEA, 2006; TMEA, 2006). School Selection Procedures The researcher contacted representatives of the vocal associations and bandmasters associations in Alabama, Florida, Georgia, and Tennessee by email and telephone to identify band and choral programs with the greatest number of students participating in all-state festivals in 2007 and 2008. The AVA provided copies of the 2008 concert programs listing all-state participants (P. Edmundson, personal communication, March 25, 2008). The ABA (G. Gooch, personal communication, April 3, 2008), TBA (Z. Williamson, personal communication, April 2, 2008), and the band division of GMEA (G. Gribble, personal communication, March 24, 2008) provided lists of all-state band participants. The researcher purchased a list of vocal all-state festival participants from the FVA (E. McNamara, personal communication, March 31, 2008). Lists of vocal all-state students were retrieved from the Tennessee vocal associations? web sites (ETVA, 2008; MTVA, 2008; WTVA, 2008). Schools that had both choir and band all-state members were selected to eliminate effects of differing school emphases on chorus or band programs. It seemed logical that schools with participation in both performance areas were more likely to 52 have a balance of emphasis. Factors were not as likely to be skewed by the emphasis of the school. Fourteen Alabama schools, 14 Georgia schools, and 4 Tennessee schools fit the criterion for the selection. Distribution of all state choir members across Florida schools differed from the distribution in Alabama, Georgia, and Tennessee. Four Florida schools had the highest number of vocal participants in the state: three participants. Twenty-two schools had two vocal participants and the remaining schools had only one participant per school. The limited number of students represented in each school resulted in a broad representation of students from the state, but did not insure that the best vocalists in the state were represented. The distribution of participants from each school in Alabama, Georgia, and Tennessee indicated that the best students were concentrated in a smaller number of schools. Florida was not included in the study for two reasons. First, the distribution of all-state students in FVA did not represent performance success as accurately as Alabama, Georgia, and Tennessee. Secondly, the collection of data that would include a significant number of all-state participants was impractical. Access and Permission The study protocol was approved by the Auburn University Institutional Review Board?s expedited procedure. The researcher contacted prospective principals or superintendents of 25 high schools to request their participation in the study. The researcher explained the procedures used to guarantee anonymity and confidentiality and provided each principal, teacher, and parent/guardian with contact information for the researcher and the Auburn University Office of Human Subjects. The researcher 53 initially contacted each principal by telephone to describe the study and ask permission to administer the survey in their school. If the principal was not authorized by the school system to approve the research, the principal referred the researcher to the appropriate administrator. Authorizations were provided by superintendents, assistant superintendents, accountability specialists, evaluation specialists, and fine arts coordinators. (See Appendices C and D for human subjects approval and recruiting materials.) The researcher sent a letter by email to the authorizing school representatives requesting written permission to administer the survey. A sample consent letter, to be returned to the Office of Human Subjects Research at Auburn University, was attached to the email. After the first two contacts, the researcher determined that it was more efficient to indicate that a report would be emailed to all administrators and teachers who participated instead of having a request form for the report. Upon receipt of the authorizing representative?s verbal permission, the researcher contacted each band and choral director by phone to describe the study, request their consent to administer the survey to their classes, and obtain student enrollment. After confirmation from the Office of Human Subjects Research the researcher mailed each director a packet containing a letter with instructions, two parent permission forms per student, and one questionnaire per student. Data Collection Procedures Each teacher was asked to administer the Characteristics of High School Music Students Survey (CHSMSS) to all students who had returned a permission form 54 during a band or choir class. The teacher or an appointed student collected the completed surveys and returned them with the permission forms in a self-addressed, postpaid envelope. Instrument Two previously developed instruments, AMF (Asmus 1985a, 1989) and PIM (Zdzinski 1992, 1993) were combined to create the Characteristics of High School Music Students Survey (CHSMSS) (see Appendix A) . Evaluation of Pilot Study A pilot study was designed to evaluate the data collection procedures and the survey instrument. Two hundred music students at Auburn High School were recruited for a pilot study of validity and reliability of the instrument and procedures. An Exploratory Factor Analysis (EFA) was used to identify factors related to student participation in band and chorus. The resulting sample size of 80 (13 choral students and 68 band students) was too small for factor analysis. The KMO Measure of Sampling Adequacy Test (KMO = .21) was below .70 and indicated inadequate correlations to proceed with factor analysis. The KMO result was expected because of the small sample size. A minimum of 642 participants was necessary to perform a factor analysis for the 74 items in the survey (Meyers, Gamst, & Guarino, 2006, p. 567). The band students completed the 89-item CHSMSS survey (see Appendix B) during class time and the choral students completed the survey at home. Twenty percent (n = 14) of the group that took the survey home and 55% of the group that 55 took the survey in class returned completed questionnaires. When the band director and choral director were asked their opinions concerning taking the survey in class, both affirmed that it should be taken in class. One chorus student was excluded from the study, because no effort to respond correctly was apparent. One band student was excluded from the study, because the student did not respond to most of the items. The ages of the participants ranged from 14 - 18 (M = 16.25, SD = 1.0). One student was 14 years old and nine students were 18 years old. The participants were evenly distributed across grades 10 - 12: 28 sophomores, 27 juniors, and 25 seniors. Most of the students lived with both parents (81.2%). No distinction was made between parents and stepparents in the number of parents living at home. Only one student reported living with a guardian with no parents at home. The mean number of siblings in the families of the participants was 1.50, which included siblings living at home and those not living at home. The average family occupational level, educational level, and SES are shown in Table 3. The average SES of the families in the pilot study ranks in highest category of Hollingshead?s (1958) index of social position. The highest category is defined by a score of 64-77. Nearly half the families (41.2%) scored 77 on the Hollingshead index and 74.9 % of the families scored 64 or higher. 56 Table 3 SES of Pilot Study Participants M SD Occupational Level 6.29 1.38 Educational Level 6.62 .60 SES 68.26 11.23 Six items were deleted from the survey as a result of the pilot study. Item 76, ?years of private lessons,? and item 77, ?amount of practice time,? did not converge into any of the factor components. The two items did not sufficiently relate to the research questions so they were removed from the CHSMSS. In the dataset, item 78 (all-state chorus) and item 79 (all-state band) were combined into a categorical variable identifying the student as a band participant or choir participant. Item 80 identified the number of parents living at home. If only one parent lived at home, the data coding was different for father and mother. By using different coding for each parent, the researcher could identify whether there was any difference in performance achievement, parental involvement, or motivating factors between fathers and mothers in single parent homes. After reviewing the results, the researcher determined that the questions about parents living at home (item 80) and number of siblings (item 81) did not add any benefit to the survey beyond the items from Zdzinski?s (1992, 1993) PIM. Items 80 and 81 were deleted from the demographic section. Data from item 84 (number of brothers) and item 85 (number of sisters) were coded in one cell as number 57 of siblings, but the items were deleted from the CHSMSS, because they did not sufficiently relate to the research questions. The deletions left only items that were included in the Asmus (1985a) scale and the Zdzinski (1992, 1993) scale to be analyzed. Some of the responses to the questions about parents? occupations were answered with descriptions that were too general to categorize so the researcher added, ?Please be as specific as possible? to those questions. Structure of the CHSMSS Items 1 through 35 on the CHSMSS were designed with the same wording and formatting as the AMF to measure students? motivation factors through attributions of success and failure in music. The responses were coded into SPSS on a scale of 1 (?not important at all?) to 5 (?extremely important?). Items 36 through 74 were identical to PIM items and measured family background and parental involvement. Items 36 through 50 were from the PI-F subscale. If the student?s response to an item was A, indicating that the parent(s) were always involved, the item was coded as 5. A response of E, indicating that the parent(s) were never involved, was coded as 1. Items 51 through 74 represented the PI-D subscale. If the student indicated that neither parent was involved in items 51 through 65, the item was coded as 0. If the father only or mother only was involved, the item was coded as 1. If both parents were involved, the item was coded as 2. Items 67 through 73 measured parental involvement in creating a home music environment through yes (coded 1) or no (coded 0) responses. Item 74 measured a degree of parental involvement for band students based on instrument ownership. School owned instruments were coded as 1, rented instruments 58 were coded as 2, and family owned instruments were coded 3. Students indicated their membership in band or chorus in item 75. In order to maintain confidentiality in the surveys and to minimize threats to internal validity, teacher recommendation was not used as the criteria for identifying students who are successful in band or choral performance achievement. Selection validity (Pedhazur & Schmelkin 1991) would have been jeopardized because variances in teachers? criteria for success would create inconsistencies in the success group. If the teacher identified the students before the survey was administered, the students? responses would have been subject to compensatory rivalry or resentful demoralization (Pedhazur & Schmelkin 1991). If the teacher identified the students after the survey was administered implementation of the survey could not have been anonymous and the participants might not have felt free to express their opinions. The criteria used to establish performance achievement was the students? participation in all-state band or chorus. In Alabama, Georgia, and Tennessee, expert judges assessed the band students? according to similar measurements of success and the choral students according to similar measurements of success. Students? performance achievement in chorus or band was identified by all-state chorus or band participation in item 76. Students responded to item 76 by listing the number of years they had participated in all-state chorus or band. The response to all-state participation was treated as a dichotomous variable. Any response of one or more years was identified as successful achievement in performance. Participation was operationalized by labeling participation of one or more years as 2 and non-participation as 1. 59 The 89-item version of the CHSMSS used for the pilot study differed slightly from the final 83-item survey as was explained in the description of the pilot study. The last portion of the CHSMSS identified demographic information about the students. Students were asked to indicate their gender, age, and grade level. The survey contained four questions to identify SES concerning each parent?s occupation and education. The format of the occupation and education items was patterned after Dunlap?s (1975) survey. SES was classified using the Hollingshead Index of Social Position (Hollingshead & Redlich, 1958). The researcher took steps to identify students who might have been surveyed twice. The researcher numbered each survey in a band or choral group within a specific school so that the subject number in SPSS matched the survey number written on the survey. The researcher catalogued the survey numbers for each group. None of the participants indicate that they participated in band and chorus, so no further steps were warranted to eliminate duplicates. Data Analysis The data analysis was designed to address the purposes and research questions of the study. The primary purpose of this study was to identify the contributions of parental involvement, motivating factors (attributions of success), and SES to performance achievement among high school music ensemble members. The secondary purpose of this study was to compare parental involvement, motivating factors, and SES of high school choir and high school band members. Parental involvement was measured through the PIM scale. The motivating factors were 60 identified through factor analysis of the AMF scale. Hollingshead?s Two Factor Index of Social Status was used to measure SES. The research questions were stated as follows: 1. What are the parental support factors, motivational factors, and SES of high school band and choir students who attend high schools that have both choir and band students selected for all-state participation? 2. How do the factors relate to membership in band or choir ensembles? 3. How do the factors relate to all-state participation in band and choir students? Data was entered into SPSS (16.0) statistical software. Research question 1 was addressed through reliability analyses, CFAs, and EFAs. Internal consistency reliability was established for each factor identified in the PIM scale, the three PIM subscales in the PIM, and the AMF scale. Parental involvement was determined through the PIM sub-scale scores. The PI-F range was 15 ? 75, the PI-D range was 0- 30, and the PI-C range was 9 ? 18 (yes = 2, no = 1). The researcher ran a CFA using AMOS (16.0) software to confirm that the factor structure identified in the AMF scale (items 1 ? 35) fit the current data sample. Motivating factors related to band and chorus participation were identified through EFA and CFA of the AMF scale. A CFA of the AMF scale was designed to minimize the possibility of Type II error in items 1 - 35. The CFA measured the fit of the current data to the Asmus? (1986b, 1989) model. (Asmus had identified the factor structure through a principal components factor analysis.) 61 Two relative fit indices, the Normed Fit Index (NFI) and the Comparative Fit Index (CFI), and one absolute fit index, the Root Mean Square Error of Approximation (RMSEA), were calculated to fit the model to the data (Guarino, Shannon, & Ross, 2001). The NFI and CFI indicate the improvement of the model over the independence model, which assumes that there are no relationships within the data. The NFI and CFI are probability values that range from 0 to 1 and should be > .95 to indicate a good fit (Meyers, Gamst, & Guarino, 2006, p. 575-576). ?The RMSEA is the average of the residuals between the observed correlation/covariance from the sample and the expected model estimated from the population? and should be < .08 (Meyers, Gamst, & Guarino, 2006, p. 576) or < .06 according to Schreiber, Stage, King, Nora, and Barlow (2006). After identifying the best factor structure fit through CFAs, the researcher ran a principal components factor analysis with a varimax rotation. The factor loading of the current sample was compared to Asmus? (1989) factor loading. The KMO Measure of Sampling Adequacy Test, > .70, was used to indicate whether adequate correlations existed to proceed with factor analysis. Variables with correlations of .3 or higher were identified as part of a component. Components with eigenvalues ? 1 were reported as contributors to the total variance of the factors. In a factor analysis, the term extracting components describes the process of grouping variables into components. A component is a group of variables in a factor analysis that are highly correlated. The extracted components are identified only as component 1, component 2, etc. The components have meaning as a factor when the 62 researcher identifies the component through the similarities of variables in the component. Once the researcher has labeled the components, they are referred to as factors. An eigenvalue indicates how much of variance of the initial group of variables is accounted for by one component. An eigenvalue is the sum of the squared (r2). Research question 2 sought to identify the relationship of the factors to band and choir participation and research question 3 sought to identify the relationship of the factors to all-state participation. DFA was determined to be most appropriate analysis to identify differences between band and chorus members and all-state and non-all-state participants, because the researcher used nine continuous variables (AMF factors, PIM subscales, and SES) to predict success in band students and choral students (Asmus & Radocy, 1992, p 160; Meyers, Gamst, & Guarino, 2006, chap. 7). Pearson product-moment correlations were analyzed between the individual items of the PIM scale and all-state participation. The correlations were compared to the correlations of individual PIM items found by Zdzinski (1996). 63 CHAPTER 4 RESULTS Introduction The results of this study are organized according to the chronology of data collection, and order of the analyses used to answer each of the research questions. Three research questions are addressed through the analyses: 1. What are the parental support factors, motivational factors, and SES of high school band and choir students for all-state participation? 2. How do the factors relate to membership in band or choir ensembles? 3. How do the factors relate to all-state participation in band and choir students? The sections included in the results chapter are survey response, reliability, factor analysis of the AMF scale, comparison of means, correlation analysis of the PIM scale, comparison of factors between groups, and DFA. Response Rate and Sample Characteristics Thirty-three high schools in Alabama, Georgia, and Tennessee had at least six students participating in all-state band and six students participating in all-state chorus during the 2007-2008 academic year. Fifteen of the schools were in Alabama, 14 of the schools were in Georgia, and 4 of the schools were in Tennessee. Principals, superintendents, or arts coordinators from 11 schools agreed to conduct the research in 64 their schools. The approval process for research in two school systems in Georgia took a minimum of three months for approval. If the research were approved, it would be for the 2008-2009 academic year. The researcher did not pursue approval from those 11 schools, because the other schools that responded positively accounted for approximately 2,300 potential participants. Administrators for 9 of the 11 schools who had verbally agreed to conduct the research returned a letter of permission to the Auburn University IRB. The researcher sent 1,891 surveys to the band and choir directors of the nine schools. Five schools returned a combined 323 surveys for a return rate of 17%. These surveys and the 80 surveys collected in the pilot study represented 403 participants from six schools, resulting in a return rate of 19.27%. The minimum sample size needed for a factor analysis was 50 plus 8 multiplied by the number of variables (34), or 322 participants (Meyers, Gamst, & Guarino, 2006, p. 567). Thirty-four of the 35 variables in the AMF were analyzed through factor analysis. The researcher?s intent was to survey both band and chorus students in each school, but the returned surveys did not include students from both programs in all the schools. Only two schools had participation from both band and chorus students. The sample, however, met the objective of surveying band and choral students from schools in which both programs had significant representation at all-state festivals (see Table 4). The mean age of the 403 participants was 16.01 and the mean grade level was 10.28. The participants included 185 (45.9%) chorus participants and 218 (54.1%) band participants. One hundred forty-two participants (45.9%) were male and 257 65 (63.8%) were female. Three participants did not respond to the gender survey item. One hundred twenty-four (30.8%) of the participants indicated participation in one or more years of all-state ensembles including 58 chorus participants and 66 band participants. Table 4 Summary of Participants by State and Ensemble Type Schools Band Chorus Total P AL 4 154 89 243 60.30 GA 1 64 0 64 15.88 TN 1 0 96 96 23.82 Total 6 218 185 403 100.00 Research Question 1 Three scales were used to identify factors related to band and chorus participation. Parental involvement factors were identified through the PIM, motivational factors were identified through the AMF, and SES was measured through Hollingshead?s Index of Social Position. Reliability Reliabilities for PIM and AMF scales were estimated through calculation of Cronbach?s alpha. Reliability coefficients for the three PIM sub-scales are presented in Table 5 with comparisons to Zdzinski?s findings.PIM reliability (? = .911) was greater than Zdzinski?s ? = .848. Reliability estimates for the five factors in the AMF are 66 presented in Table 6. The AMF reliability was ? = .905 (M = 132.98, SD = 16.28), which was greater than Asmus? (1986a, 1986b, 1988) reported AMF reliability ? = .728 (M = 118.58, SD = 16.81) and Zdzinski?s (1996) ? =.866(M = 118.58, SD = 16.81). Table 5 Reliability of Zdzinski?s PIM Scale and Sub-scales Scale Zdzinski Hickok M SD ? M SD ? PIM 50.08 12.741 .848 63.09 17.172 .911 PI-F 35.57 8.726 36.56 10.800 .859 PI-D 14.30 5.313 11.52 8.067 .840 PI-C 11.98 2.009 .641 67 Table 6 Reliability of Asmus? AMF Scale Scale Asmus Zdzinski Hickok ? M SD ? M SD ? AMF .728 118.58 16.807 .866 132.98 16.277 .905 Effort 24.82 4.195 .854 Classroom 22.48 4.492 .790 Ability 26.14 3.748 .853 Background 17.33 4.918 .806 Affect 24.04 3.882 .765 Parental Involvement Parental support factors were measured through the PIM. The frequency of parental involvement, the degree of parental involvement and the home music environment were established through the three subscales of the PIM (Zdzinski, 1993). The range of the PI-D sub-scale was changed to 8 ? 16. Item 74 on the survey referred to instrument ownership. The item was relevant to band students, but not choral students, so it was deleted. Analysis of Motivating Factors The fit of Asmus? factor model to the current data was identified through a CFA. The NFI in Asmus? 35-item model indicated a poor fit (see Table 7). The CFI 68 indicated a slightly better fit, but still poor. The RMSEA was .070 indicating a good fit of the model to the data according to the criteria of < .08 suggested by Meyers, Gamst, and Guarino (2006, p. 574). The RMSEA, however, did not meet the criteria of < .06 recommended by Schreiber et al. (2006). Although the preliminary fit indices indicated a poor factor structure, the internal consistency reliability indicated that the AMF factor structure was appropriate for the current study. The CFA fit indices were slightly better when item 27 ?afford a good instrument? was deleted from the Asmus model. The researcher determined that although item 27 was relevant to band participants, it was not relevant to choral participants. Further explanation concerning the researcher?s decision to use the 34-item model is included in the following section on EFAs. Table7 shows a comparison of the fit indices of the model with 35 variables to the model with 34 variables. The fit indices of the 34-variable model are reported for all participants, band students only, and chorus students only. 69 Table 7 Fit Indices of the Asmus Model and Study Data Fit Index All participants 35 items All participants 34 items Band 34 items Chorus 34 items NFI .729 .739 .662 .635 CFI .799 .807 .763 .756 RMSEA .070 .070 .081 .080 ?2 2.991* 2.984* 2.436* 2.176* * = p < .01 Chi-square tests were calculated to compare the 34-item and the 35-item models (Thompson, 2004). The fit of the model was slightly better again when all 8 of the variables that did not load into the original model were omitted. The NFI, CFI, and RMSEA fit indices were similar for the band participants and chorus participants, but were not as good as in indices for all participants. The chi-square value was better for the band and chorus models than for the model with all participants. In all cases the model ?2 was significant (p < .01). EFAs were used to determine variances in the factor structure that could improve the fit of the model for the current study. Nine variables in the current data set loaded onto components other than the original Asmus model with loading coefficients greater than .3 in a second component. As seen in Appendix E, in six of the nine cases, the second component was the same factor in which the variable loaded in Asmus? model. The three variables that loaded into 70 component 6 were variables that Asmus identified as background variables. The variable that did not indicate any relationship to the original factor structure was ?being able to afford a good instrument? (item 27 of the survey). It is reasonable to assume that the deletion of the nine variables that did not load as expected would create a better fit of the model. Variables with poor fit can be deleted from the model until only variables that fit the model well are retained. Although the model can be made to fit, valuable data would be omitted. The same factors were identified in EFAs with and without the nine variables. The omission of item 27 did not change the factors that were identified, but, it changed the amount of variance explained by each factor. Item 27 was relevant to Asmus? study of instrumental music participation, but was not relevant to the study of choral music participation; therefore, it was deleted from the data set. An EFA of all participants and an EFA of only band students, without item 27, yielded seven components. All of the variables in the AMF had a loading coefficient greater than .3 in the seven components (all with eigenvalues greater than 1) of the current model. An EFA of choral students yielded eight components. The KMO measure of sampling adequacy in each model was > .70, so the ratio of subjects to variables was good. For all participants in the 35-item scale the KMO was .889. In the 34-item model the KMO was .888 for all participants, .854 for band students, and .823 for choral students. In Asmus? study and in the current study Effort had the highest eigenvalue. The order of the remaining four factors in Asmus? study was Class Environment, Musical Ability, Background, and Affect for Music. The table shown in 71 Appendix F illustrates the rotated component matrix without the variable ?able to afford a good instrument.? All loading coefficients of .3 or greater are included in the table. Five variables had a loading coefficient of .3 or greater in more than one factor. Seven of the variables did not load under the expected factor of the Asmus model. Four of the variables that did not load with the expected factor had a loading factor of .3 or greater in a second factor that corresponded with the Asmus model. The first five components, which account for 49.58% of the variance, were given factor labels consistent with the Asmus factors. Components six and seven were not named. The researcher analyzed the data through a third EFA in which the number of factors was suppressed. When suppressing the analysis to five factors, all but two of the variables loaded onto the same factor as they did in Asmus? (1986b) study. As seen in Appendix G, both of the factors also had a factor loading < .3 in the expected factor. A similar EFA was done with band participants suppressed to five factors. The items loaded onto the same factors with band students as they did with all participants, but the variance explained by each factor was different. When the same analysis was done with chorus students only, most of the effort and ability variables were loaded onto the first component. The fifth component included two variables related to musical ability and one variable related to affect for music. Table 8 illustrates the variance explained (rotation sums of squared loadings) by the factors in the current sample with all participants, band students, and chorus students. In Asmus? study, the factors in order of explained variance are Effort, Classroom Environment, Musical Ability, 72 Background, and Affect for Music. The fifth component in the suppressed analysis of chorus participants did not include enough variables to clearly identify it as a factor, so it is identified only as a component. The EFAs that were not suppressed contain more than five components, but only the components identified as factors based on the Asmus scale are reported. Item 27 (?able to afford a good instrument?) is not included in the current data set. Table 8 Percentages of Variance Explained in the 34-Item AMF Scale Factors All All Suppressed Band Band Suppressed Chorus Chorus Suppressed 1 12.04 Ef 11.59 Ab 11.87 Ef 11.90 Ef 14.35 Ab 18.43 Ef/Ab 2 10.51 Ab 11.59 Ef 11.06 Ab 11.82 Ab 9.26 Ef 11.88 Bk 3 9.57 Af 10.13 Bk 11.01 Af 11.61 Af 8.70 Af 8.91 Cl 4 9.00 Cl 10.02 Af 10.45 Cl 10.46 Cl 7.68 Cl 8.52 Af 5 8.46 Bk 9.70 Cl 7.48 Bk 9.62 Bk 7.51 Bk 5.76 Co 6 6.84 Bk Total 49.5 53.02 51.87 55.40 54.34 53.51 Note: Ef = Effort, Cl = Classroom Environment, Ab = Ability, Bk = Background, Af = Affect for Music, Co = Component The EFAs with all participants and the EFA with band participants yielded seven components. The EFA with chorus participants resulted in eight components, 73 but components five and six both included background variables. Although the order of the proportion of explained variance differed between all participants and band participants, the difference in variance of like factors ranged from .45% - 1.67%. The greatest difference in the percentage of variance for all factors was 3.58% for all participants and 1.89% for band participants. The factor loadings and variances of the factors in the EFA for choral students were not consistent with the factor loadings and variances of band students and all participants. These differences support the need to further analyze the differences between factors related to band participation and chorus participation. The CFAs indicated a minimal fit of the factor model to the current data but the EFAs resulted in factor loadings that were similar to those identified by Asmus. In the EFA of band participants, when suppressed to five factors, all variables loaded to the same factors as they did in the Asmus (1989) study. With the exception of two variables, the EFA of all participants resulted in the same match. The match of factor loadings between the current data and the Asmus study implies that Asmus? factor model can be applied to all participants in the current study. Factors Related to Band, Chorus, and All-state Participation Nine factors were established to identify the relationship of parental involvement, student attributions of success, and SES to all-state participation and band or chorus participation. Three factors which were established through the PIM subscales are frequency of parental involvement (PI-F), degree of parental involvement (PI-D), and parental provisions for the home musical environment (PI-C). 74 Five factors which were a result of the factor analysis of the AMF scale were Effort, Background, Classroom Environment, Musical Ability, and Affect for Music. SES was established through Hollingshead?s (1958) Occupational and Educational Scale. (See Table 9 for descriptive statistics for each factor.) Table 9 Descriptive Statistics for CHSMSS Factor Means Factor N Minimum Possible Maximum Possible M SD PI-F 403 15 75 36.56 10.800 PI-D 403 0 30 11.52 8.067 PI-C 403 8 16 11.98 2.009 Musical Ability 403 7 35 26.14 3.748 Effort 403 7 35 24.82 4.195 Affect for Music 403 7 35 24.04 3.882 Classroom Environment 403 7 35 22.48 4.492 Background 403 6 30 17.33 4.918 SES 401 11 77 59.40 17.027 Research Question 2 The DFA for band and chorus membership was computed to address research question 2: How do the factors relate to membership in band and choir ensembles? Classification results of the DFA indicated that 63.3% of the participants were 75 correctly classified as band or chorus participants according to predictions based on the CHSMSS scores. Predictions based on the CHSMSS scores would have been 71.0% correct for the students currently in band and 54.3% correct for students currently in chorus. The overall multivariate function was statistically significant for band and choir membership (Wilks?s Lambda = .887, p < .001). Table 10 illustrates the difference in the means of each factor according to band or chorus membership. The follow up F tests revealed significant main effects for Ability and SES with band or chorus participation. The mean score for the Ability in Music attribution factor was higher for band students than for chorus students. The mean SES was higher for chorus students than for band students. 76 Table 10 DFA Means for Band and Chorus Participation Band Chorus Factor M SD M SD F PI-F 36.41 9.887 36.97 11.643 .269 PI-D 11.78 8.132 11.33 7.958 .321 PI-C 11.90 2.024 12.11 1.958 1.057 Ability 26.80* 17.911* 25.39* 4.116* 14.589* Effort 24.85 3.251 24.78 4.167 .032 Affect 23.77 4.020 24.36 3.676 2.276 Class 22.16 4.542 22.84 4.395 2.270 Background 16.99 4.673 17.68 5.120 2.006 SES 54.87* 17.911* 62.39* 15.441* 10.739* * = p < .001 Research Question 3 The results of the DFA for all-state and non-all-state participation address research question 3: How do the factors relate to all-state participation in band and choir students? The results indicated that 71.3% of the participants would have been correctly identified as all-state or non-all-state participants. Predictions would have been 95.3% correct for the students who did not participate in all-state ensembles, but only 17.1% correct for students who were all state participants. The overall 77 multivariate function was statistically significant for all-state and non-all-state participation (Wilks?s Lambda = .914, p < .001). Table 11 illustrates the difference in the means of each factor according to all-state participation and non-all-state participation. The follow-up F tests for all-state and non-all-state participation revealed a significant main effect for all factors related to Background and parental involvement. Background, PI-F, PI-D, and PI-C were higher for all-state participants than for non-all-state participants. The strongest effect was the PI-F score. The results of the DFAs suggest that four factors that have predictive value for all-state participation: Background, Frequency of Parental Involvement, Degree of Parental Involvement, and Parental Provisions for the home musical environment (see Table 11). 78 Table 11 DFA Means for All-state and Non-all-state Participation All-state Non -all-state Factor M SD M SD F SES 60.85 15.969 58.76 17.464 1.291 Ability 26.29 3.566 26.09 3.812 .251 Effort 25.19 4.007 24.65 4.270 1.375 Affect 24.45 3.794 23.86 3.899 1.944 Class 22.43 4.705 22.49 4.389 .014 Background 18.33 ** 4.692 ** 16.85 ** 4.914 ** 7.997 ** PI-F 39.84 *** 11.271 *** 35.26 *** 10.173 *** 16.122 *** PI-D 13.53 *** 8.492 *** 10.71 *** 7.698 *** 10.729 *** PI-C 12.40 * 1.867 * 11.82 * 2.026 * 7.281 * * = p ? .007; ** = p ? .005 level; *** = p ? .001 Correlations of Variables in the PIM A correlation analysis was used as a follow-up test to compare individual items on the PIM to the correlations that Zdzinski (1996) reported. Because Zdzinski used Pearson product-moment correlations to identify relationships between the PIM items and performance ability, the researcher used Pearson Product-Moment Correlations to measure relationships among PIM variables and all-state participation (Shannon & 79 Davenport, 2001). Although Zdzinski?s measures differed from the all-state measurement in the current study, both studies used valid measurements for performance ability. The comparison of the two correlations related to the parental involvement aspect of research question 3. Table 12 illustrates the comparisons of Zdzinski?s study and the current study. Correlations that are missing in Zdzinski?s list are a result of the reporting method and are not statistically significant. Zdzinski measured performance ability, affective musical ability, and cognitive musical ability and reported only PIM items that had a significant correlation with at least one of the musical ability assessments. Table 12 Pearson Product-Moment Correlations for Variables in the PIM Variable All-state Membership Zdzinski Performance Assessment PI-F items Attend parent meetings .24 ** .18 ** Talk about music .23 ** .10 Attend school concerts .21 ** .30 ** Attend non-school concerts .17 ** .21 ** Listen to practice .15 ** -.15 Ask about progress .14 ** -.12 Record performances .12 * .19 ** 80 Table 12 (continued) Variable All-state Membership Zdzinski Performance Assessment Sing with you .09 .00 Attend school rehearsals .09 Assist with practice -.09 -.30 Take you to concerts .06 .02 Listen to music at home -.05 .06 Play in group .04 .08 Provide transportation -.04 .01 Sing in group -.01 .17 ** PI-D items Music parent organization .37 ** .22 ** Attend parent meetings .18 ** .14 ** Talk about music .18 ** .12 Attend school concerts .18 ** .20 ** Assist with practice .14 ** -.26 ** Attend non-school concerts .14 ** .21 ** Ask about progress .12 * .06 Listen to practice .10 -.09 Record performances .09 .12 81 Table 12 (continued) Variable All-state Membership Zdzinski Performance Assessment Play in music group .06 Take you to concerts .04 .07 Attend rehearsals -.05 Sing in music group .04 .12 Provide transportation .04 .19 ** Listen at home -.00 .06 PI-C items Own classical recordings .18 ** .06 Siblings sing or play .12 * .25 ** Provide recordings .08 .26 ** Purchase music .08 .19 ** Take lessons .07 Provide toy instruments .06 .10 Give you lessons .05 Play or sing with you .03 * = p ? .05; ** = p ? .01 82 Twelve of the 17 PIM items that correlated with performance measurements in Zdzinski?s (1996) study correlated with all-state participation in the current study. Five items in the current study that correlated with all-state participation did not correlate with performance measurements in Zdzinski?s study. The results illustrate similarities in the relationships of the items to the factors in the two studies. 83 CHAPTER 5 DISCUSSION Response to Research Questions The first research question was, ?What are the parental support factors, motivational factors, and SES of high school band and choir students who attend high schools that have both choir and band students selected for all-state participation?? The results of the DFA for the PIM support Zdzinski?s (1993) findings that parental involvement is significantly related to performance outcomes in instrumental students. The reliability of the PIM scale supports the identification of degree and frequency of parental involvement and the home musical environment as factors related to band and chorus participation. The results of the EFA for the AMF scale support Asmus? (1986) identification of Effort, Background, Classroom Environment, Musical Ability, and Affect for Music as attributions related to success in instrumental students. The results of the EFA for band and choral students indicate that the same attributions are related to success in choral students. The mean SES of all participants was in the second highest of the five Hollingshead classes of social position. The second research question was, ?How do these factors relate to membership in band or choir ensembles?? Band students appear to perceive musical ability as more important than chorus students do. The mean SES is higher for chorus students than 84 for band students. The average SES for chorus students was in the highest class of social position identified by Hollingshead & Redlich (1958). The third research question was, ?How do the factors relate to all-state participation in band and chorus?? The students? perceived importance of background, the frequency of parental involvement, the degree of parental involvement, and the home musical environment were higher for all-state students than for non-all-state students. The item ?caring about music? appears to have been interpreted differently by many of the participants in the current study, than it was in Asmus? study. In the Asmus study, the item was related to Effort, but in the current study the item was related to Affect for Music in all the factor analyses except the analysis of all-state participants. It appears that those who are most successful in music performance interpret ?caring about music? as an important part of Effort. A DFA identified musical ability attributions and SES as the factors that best described the difference between band and chorus students. Chorus students reported higher SES scores and band students reported higher attributions of musical ability. One might rationalize that family environment and parental involvement could cause students to choose band or chorus, but such a supposition cannot be established about a student attribution. There is no indication whether a student?s attribution of the importance of musical ability is established before choosing band or chorus participation. It is possible that a student?s perception of the importance of musical ability is formed as a result of participation in band or chorus. 85 Predicting Success in Band and Chorus The CHSMSS appears to be of moderate value in deciding whether to participate in band or in chorus. It is not clear whether the band or chorus experience develops the attributions of success or whether attributions affect the students? decisions toward participation in band or chorus. The classification results of the DFA do not indicate that the composite CHSMSS is an accurate indicator in predicting students? participation in band or chorus. Even though the predictive value of the CHSMSS was high for non-all-state participants, the CHSMSS did not appear to be a strong predictor for students who participated in all state. The composite CHSMSS does not appear to be a strong predictor of success in performance ability, but portions of the scale appear to have some predictive value. The strong correlation between all- state participation and the parental involvement and home environment factors imply that higher levels of parental involvement can contribute to a student?s participation in all-state. Implications for Parents The parental involvement and family environment factors appear to have the most significant effect on students? success in music performance as measured by all- state participation. Children perceive family background as an important part of their musical endeavors. Parents must provide their children resources for success in music. The number of parents involved with their children?s musical endeavors was significant, but the most important significant effect of all the parental involvement and family environment factors was the frequency of involvement by parents. The 86 results indicate that effective parental support can be measured tangibly by the time that parents commit to their child?s musical education. Implications for Educators One might reasonably believe that student attributions in music are affected by their participation in band and chorus. A positive or negative environment within a classroom could affect a student?s perception of the importance of the classroom environment to one?s success in music. Students who do not consider themselves successful in music might attribute their lack of success to a lack of musical ability rather than a lack of effort. Students who have put forth considerable effort and found success in music could attribute their success to effort, even if their musical ability played a significant role in their success. If Hallam and Shaw?s (2002) findings that musical ability is learned rather than innate were correct, they would emphasize the importance of effort. Educators can be encouraged to know that within the attribution factors, students consistently identified effort as the first or second most important attribution for success. It is important for teachers to continue to emphasize the importance of effort to their students. The classroom environment can be a contributing factor to the students? perceptions of the importance of effort. Band and chorus teachers have opportunities to communicate the effects of parental involvement through their parent organizations. The effect of parental involvement on all-state participation underscores the importance of maintaining a strong parent organization. Recruiting parents to be involved and explaining the 87 effects of that involvement can enhance the students? opportunities for success in band and chorus. Recommendations for Further Research Further study is warranted on the relationship of the attribution and parental involvement factors to choral students on a wider spectrum of SES than is included in the current study. The current study identified schools with a significant number of all- state participation in band and chorus. One can reasonably assume that these are strong band and chorus programs. Schools with strong programs in both areas might not represent the lowest SES population of Alabama, Georgia, and Tennessee. The classroom environment is likely to be more positive in the schools represented in this study than in many schools that struggle financially and have a lower SES population. Most of the existing studies have investigated music students currently participating in music performance ensembles. An investigation of students who have never participated in music programs in addition to those who currently participate might provide a clearer assessment of the attributions and parental involvement factors that relate to success in music performance. Further research is warranted into students who previously participated in music programs, but dropped out. What are the reasons they dropped out? Pitts, Davidson, and McPherson?s (2000) found that persistence in instrumental students was related to parental involvement. The role of parental involvement might be similar for persistence in choral students. Differences between students who currently participate in band and chorus and those who have dropped out are likely to be similar to differences between students who demonstrate performance 88 achievement and those who do not. Identifying variances in student attributions of success, family environment and parental participation would help educators to identify and avoid factors that lead to failure in music performance. 89 REFERENCES Alabama Music Educators Association (2007). 2007-2008 ABA all-state audition requirements. Retrieved October 27, 2007 from http://www. alabamamea.org/bandmasters/index.htm Alabama Vocal Association (2006). Procedure for all-state auditions. Retrieved October 27, 2007, from http://www.alavocal.org/events/allstate.php Albert, D. J. (2006). Socioeconomic status and instrumental music: What does the research say about the relationship and its implications? Update: Applications of Research in Music Education, 25(1), 39-45. Arnold, J. A. (1997). A comparison of attributions for success and failure in instrumental music among 6th-, 8th-, and 10-grade students. Update, 15(2), 19-23. Asmus, E. P. (1985a). The development of a multidimensional instrument for the measurement of affective responses to music. Psychology of Music, 13, 19-30. Asmus, E. P. (1985b). Sixth graders achievement motivation: Their views of success and failure in music. Bulletin of the Council for Research in Music Education, 85, 1-13. 90 Asmus, E. P. (1986a). Factors students believe to be the causes of success or failure in music. Paper presented at the National Biennial In-service Conference of the Music Educators National Conference, Anaheim, CA. Asmus, E. P. (1986b). Student beliefs about the causes of success and failure in music: A study of achievement motivation. Journal of Research in Music Education, 34(4), 262-278. Asmus, E. P. (1988, July). The effect of music teachers on students? motivation to achieve in music. Paper presented at the Twelfth International Research Seminar, Melbourne, Australia. Asmus, E. P. (1989, February). Criterion and construct validity of two measures of motivation for music. Paper presented at the Northwest Divisional Meeting of the Music Educators National Conference, Boise, ID Asmus, E. P., & Harrison, C. S. (1994). Effects of musical aptitude, academic ability, music experience, and motivation on aural skills. Journal of Research in Music Education, 42(2), 131-144. Asmus, E. P., & Radocy, R. E. (1992). Quantitative analysis. In R. Colwell (Ed.), Handbook of research on music teaching and learning: A project of the music educators? national conference. New York: Schirmer Books. Bonifati, L. M. (1997). The impact of the home environment on success in instrumental music (Doctoral dissertation, Columbia University, 1997). Dissertation Abstracts International, 59(09), 3452. 91 Boyle, J. D. (1992). Evaluation of music ability. In R. Colwell (Ed.), Handbook of research on music teaching and learning.. New York: Schirmer Books. Boyle, J. D., & Radocy, R. E. (1986). Measurement and evaluation of musical experiences. New York: Schirmer Books. Brand, M. (1985). Development and validation of the home musical environmental scale for use at the early elementary level. Psychology of Music, 13(1), 40-48. Brand, M. (1986). Relationship between home musical environment and selected musical attributes of second-grade children. Journal of Research in Music Education, 34(2), 111-120. Br?ndstr?m, S. (2000). For whom is music education intended? Bulletin of the Council for Research in Music Education, 147, 36-39. Colwell, R. (Ed.). (1992). Handbook of research on music teaching and learning. New York: Schirmer Books. Davidson, J. W., & Borthwick, S. J. (2002). Family dynamics and family scripts: a case study of musical development. Psychology of Music, 30, 121-136. Davidson, J. W., Howe, M. J. A., Moore, D. G., & Sloboda, J. A. (1996). The role of parental influences in the development of musical performance. British Journal of Developmental Psychology, 14, 199-412. Davis-Kean, P. E. (2005). The influence of parent education and family income on child achievement: the indirect role of parental expectations and the home environment. Journal of Family Psychology. 19(2), 294-304. 92 Deonandan, R., Campbell, K., Ostbye, T., Tummon, I., & Robertson, J. (2000). A comparison of methods for measuring socioeconomic status by occupation or postal area. Chronic Diseases in Canada, 21(3), 114-118. Dewey, J. (1933). How we think, a restatement of the relation of reflective thinking to the educative process. Boston: D. C. Heath and Company. Doan, G. R. (1973). An investigation of the relationships between parental involvement and the performance ability of violin students. (Doctoral Dissertation, Ohio State University, 1973). Dissertation Abstracts International 34(08), 5226. Dunlap, J. E. (1975). The relationship of musical achievement as measured by the Colwell Music Achievement Test to socioeconomic status, race, community size, and the presence of the father in the home in seventh-grade general music classes in Arkansas and Mississippi (Doctoral Dissertation, Indiana University, 1975). Dissertation Abstracts International, 36(05), 2477. Elliott, D. J. (1995). Music matters: A new philosophy of music education. New York: Oxford University Press. Ellis, R., Lane, W., & Olesen, V. (1963). The index of class position: An improved intercommunity measure of stratification. American Sociological Review, 28, 271-277. East Tennessee Vocal Association. (2007). 2008 all-state members. Retrieved March 21, 2008, from http://www.etva.org/Events/All_State/All_State_Roster.htm 93 Fitzpatrick, K. R. (2006). The effect of instrumental music participation and socioeconomic status on Ohio fourth-, sixth-, and ninth-grade proficiency test performance. Journal of Research in Music Education 54(1), 73-84. Florida Bandmasters Association. (2007). 2008 all-state audition requirements. Retreived October 27, 2007 from http://www.flmusiced.org/fba Frakes, L. (1984). Differences in music achievement, academic achievement, and attitude among participants, dropouts, and nonparticipants in secondary school music (Doctoral dissertation, University of Iowa, 1984). Dissertation Abstracts International (46)2, 370. Georgia Music Educators Association. (2006). District and regional audition general information and rules. Retrieved October 27, 2007 from http://www.gmea.org/info/Handbook/Band/GMEABandHdbk0607.pdf Georgia Music Educators Association. (2007). 2007-2008 GMEA handbook. Retreived October 27, 2007 from http://www.gmea.org/info/Handbook/Choral/0708ChoralHandbook0708.pdf Gordon, E. E. (1967). A three-year longitudinal predictive validity study of the Musical Aptitude Profile. Iowa City: University of Iowa Press. Gordon, E. E. (1986). Final results of a two-year longitudinal predictive validity study of the Instrument Timbre Preference Test and the Musical Aptitude Profile. Bulletin of the Council for Research in Music Education,89, 8-17. Gordon, E. E. (1998). Introduction to research and the psychology of music. Chicago: G.I.A Publications. 94 Goyder, J., & Frank, K. (2007). A scale of occupational prestige in Canada, based on NOC major groups. Canadian Journal of Sociology, 32(1), 63-83. Greenberg, M. (1970). Musical achievement and the self-concept. Journal of Research in Music Education, 18(1), 16-24. Guarino, A. J., Shannon, D. M., and Ross, M. (2001). Making sense of fit indices in structural equation modeling (SEM). Journal of Research in Education, 11(1), 130-134. Hallam, S. (1998). The predictors of achievement and dropout in instrumental tuition. Psychology of Music, 26(2), 116-132. Hallam, S., & Shaw, J. (2002). Constructions of musical ability. Bulletin of the Council for Research in Music Education, 153(4),102-108. Helwig, C., & Thomas, M. S. (1973). Predicting choral achievement through use of musicality and intelligence scores. Journal of Research in Music Education, 21(4), 276-280. Hollingshead, A. B., & Redlich, F. C. (1958). Social class and mental illness. New York: John Wiley and Sons, Inc. Hufstader, R. A. (1974). Predicting success in beginning instrumental music through use of selected tests. Journal of Research in Music Education, 22(1), 52-57. Hylton, J. B. (1981). Dimensionality in high school student participants? perceptions of the meaning of choral singing experience. Journal of Research in Music Education, 29(4), 287-303. 95 Klinedinst, R. E. (1991). Predicting performance achievement and retention of fifth- grade instrumental students. Journal of Research in Music Education, 39(3), 225-238. Kuhlman, K. (2005). Musical aptitude versus academic ability as a predictor of beginning instrumental music achievement and retention: research and implications. Update: Applications of Research in Music Education 24(1). 34- 43. Legette, R. M. (1993). Causal attributions of music majors and non-music majors regarding success and failure in music: A study of motivation and achievement. Research Perspectives in Music Education, 4, 9-12. Legette, R. M. (1998). Causal beliefs of public school students about success and failure in music (using Asmus? attribution orientation scale MAOS). Journal of Research in Music Education, 46(1), 102-111. Legette, R. M. (2003). Music achievement causal attributions as perceived by elementary public school students. Bulletin of the Council for Research in Music Education, 155(1), 44-50. Levi, R. M., Jr. (1986). The relationship of selected variables to successful choral programs (Doctoral dissertation, State University of New York at Buffalo, 1986). Dissertation Abstracts International, 47(05), 1743. (UMI No. 8609128) Lien, J. L., & Humphreys, J. T. (2001). Relationships among selected variables in the South Dakota all-state band auditions. Journal of Research in Music Education, 49(2), 146-155. 96 Manor, H. C. (1950). A study in prognosis: The guidance value of selected measures of musical aptitude, intelligence, persistence, and achievement in tonette and adaption classes for prospective instrumental students. Journal of Educational Psychology, 41, 31-50. McCarthy, J. F. (1974). The effect of individualized instruction on the performance achievement of beginning instrumentalists. Bulletin of the Council on Research in Music Education, 38, 1-16. McCarthy, J. F. (1980). Individualized instruction, student achievement, and dropout in an urban elementary instrumental music program. Journal of Research in Music Education, 28(1), 59-69. McCord, G. D. H. (2003). Choral all-state policies and practices: A survey-based analysis and critique (Doctoral dissertation, University of Georgia, 2003). Dissertation Abstracts International, 64(2). Dissertation retrieved June 3, 2008 from http://dbs.galib.uga.edu/cgi- bin/ultimate.cgi?dbs=getd&userid=galileo&serverno=8&instcode=publ Meyers, L. S., Gamst, G., & Guarino, A. J. (2006). Applied multivariate research: Design and interpretation. Thousand Oaks, CA: Sage. Middle Tennessee Vocal Association. (2006). The constitution by-laws and standing rules of the Middle Tennessee Vocal Association. Retrieved October 27, 2007 from http://www.mtva.org/2007_2008/MTVA_Handbook_9-15-07.pdf 97 Middle Tennessee Vocal Association. (2007). 2007 all-state results. Retrieved March 21, 2008, from http://www.mtva.org/2007_2008/mid_state_all_state/2007_Audition_Results. htm Miller, D. C. (Ed.). (1991). Handbook of research design and social measurement. (5th ed.). Thousand Oaks, CA: Sage Publications, Inc. Miller, D. C., & Salkind, N. J. (Eds.). (2002). Handbook of research design and social measurement. (6th ed.). Thousand Oaks, CA: Sage Publications, Inc. Mudrick, A. W. (1997). Student motivation in four successful high school choral programs in south central Pennsylvania: A qualitative study (Doctoral dissertation, Temple University, 1997). Dissertation Abstracts International, 58(10), 3869. (UMI No. 9813531) Osborn, A. F. (1987). Assessing the socioeconomic status of families. Sociology, 21(3), 429-448. Pascoe, B. J., & Waugh, R. F. (2001). The measurement of student achievement in music using a Rasch measurement. Australian Journal of Education, 45(1), 90-110. Pedhazur, E. J., & Schmelkin, L. P. (1991). Measurement, design, and analysis: An integrated approach. Hillsdale, NJ: Lawrence Erlbaum Associates. Piaget, J. (1962). Play, dreams, and imitation in the childhood (C. Gattegno & F. M. Hodgson, Trans.). New York: Norton. 98 Pitts, S. E., Davidson, J. W., & McPherson, G. D. (2000). Models of success and failure in instrumental learning: case studies of young players in the first 20 months of learning. Bulletin of the Council for Research in Music Education, 146(1), 51-69. Rainbow, E. L. (1965). A pilot study to investigate the constructs of musical aptitude. Journal of Research in Music Education, 13(1), 3-14. Rathunde, K., & Csikszentmihalyi, M. (1993). Undivided interest and the growth of talent in a longitudinal study of adolescents. Journal of Youth and Adolescence, 22(4), 385-395. Reimer, B. (2003). A philosophy of music education: Advancing the vision (3rd ed.). Upper Saddle River, NJ: Pearson Education, Inc. Schmidt, C. P. (2005). Relations among motivation, performance achievement, and music experience variables in secondary instrumental music students. Journal of Research in Music Education, 53(2), 134-147. Schreiber, J. A., Stage, F. K., King, J., Nora, A., & Barlow, E. A. (2006). Reporting structural equation modeling and confirmatory factor analysis results: A review. Journal of Educational Research, 99(6), 323-337. Shannon, D. M., & Davenport, M. A. (2001) Using SPSS to solve statistical problems: A self-instruction guide. Upper Saddle River, NJ: Prentice-Hall. Sloboda, J. A., & Howe, M. J. A. (1991). Biographical precursors of musical excellence: An interview study. Psychology of Music, 19, 3-21. 99 Slomczynski, K. M., Miller, J., & Kohn, M. L. (1981). Stratification, work, and values: A Polish-United States comparison. American Sociological Review. 46, 720-744. Tennessee Music Education Association. (2006). Tennessee Music Education Association all-state performing organizations regulations and general information 2007-2008 revision. Retreived January 27, 2008, from http://tnmea.org/All_State_Handbook.aspx Thompson, B. (2004). Exploratory and confirmatory factor analysis: Understanding concepts and applications. Washington, DC: American Psychological Association. Tipps, J. W. (2003). A preliminary study of factors that limited secondary school choral involvement of collegiate choral singers. International Journal of Research in Choral Singing, 1(1), 22-28. Tobin, N. (2005, September). A study of the music, academic, leadership and extracurricular achievements of Massachusetts all-state participants. Research and Issues in Music Education, 3(1). Retrieved October 28, 2007, from http://www.stthomas.edu/rimeonline/vol3/tobin1.htm. Van Leeuwen, M. H. D., Maas, I., & Miles, A. (2004). Creating a historical international standard classification of occupations. Historical Methods, 37(4), 186-197. Warner, L. W. (1960). Social class in America. New York: Harper Torch-books. 100 West Tennessee Vocal Association. (2005). Constitution and by-laws. Retrieved October 27, 2007 from http://www.wtvmea.org/aboutus/conby/by-laws.htm West Tennessee Vocal Music Education Association. (2007). All-state 2008 results. Retrieved March, 21, 2008, from http://www.wtvmea.org/results/index.htm Woody, R. H. (2001). Learning from the experts: Applying research in expert performance to music education. Update: Applications of Research in Music Education, 19(2), 9-14. Wright, G. K. (1996). A case study of an exemplary choral program: Issues of excellence (Doctoral dissertation, University of the Pacific, 1996). Dissertation Abstracts International, 57(04), 1536. Young, W. (1971). The role of musical aptitude, intelligence and academic achievement in predicting the musical attainment of elementary instrumental music students. Journal of Research in Music Education, 19(4), 385-398. Zdzinski, S. F. (1992). Relationships among parental involvement, music aptitude, and musical achievement of instrumental music students. Journal of Research in Music Education, 40(2), 114-125. Zdzinski, S. F. (1993). Relationships among parental involvement, selected student attributes, and learning outcomes in instrumental music (Doctoral dissertation, Indiana University, 1993). Dissertation Abstracts International, (54)12, 4386. (UMI No. 9410399). 101 Zdzinski, S. F. (1996). Parental involvement, selected student attributes, and learning outcomes in instrumental music. Journal of Research in Music Education, 44(1), 34-48. Zdzinski, S. F. (2007). The underlying structure of parental involvement-home environment in music. Unpublished manuscript. 102 APPENDIX A Characteristics of High School Music Students Survey 103 104 105 106 107 APPENDIX B Characteristics of High School Music Students Survey Pilot Version 108 109 110 111 112 113 APPENDIX C Office of Human Subjects Approval 114 115 116 117 118 119 120 121 122 123 124 APPENDIX D Communication and Permission Letters 125 126 127 128 129 130 131 132 133 APPENDIX E Rotated Component Matrix for 35 AMF Variables Variable Effort Ability Class Affect Background Comp 6 Comp 7 Practicing .740 Goals and practice .731 Practice effort .730 Trying hard .721 Serious about music .659 Willing effort .645 Feel emotion* .431 .391 Symbols and markings .807 Notes and rhythms .762 Counting music .757 Reading music .730 Sense of rhythm .567 Steady beat .462 Teacher temperament .720 Teacher favoritism .705 134 Table Appendix E (continued) Variable Effort Ability Class Affect Background Comp 6 Comp 7 Teacher understands you .656 Liking teacher .636 Liking other students .594 Getting along with others .573 Music is fun .712 Love listening .712 Caring about music* .332 .671 Like to make music .657 Please others .489 Naturally creative .450 Musical relatives .776 Runs in family .743 Musical parents .736 Being with friends* .372 .425 Natural Talent* .774 Natural ability* .749 Starting young* .411 .443 Liking sound* .476 .508 Afford a good instrument* .410 -.463 135 Table Appendix E (continued) Variable Effort Ability Class Affect Background Comp 6 Comp 7 Good ear* .363 .410 * = Items that did not load onto the factors identified in the Asmus model. 136 APPENDIX F Rotated Component Matrix for 34 AMF Variables Variable Effort Ability Affect Class Background Comp 6 Comp 7 Practicing .758 Practice effort .744 Goals and practice .726 Trying hard .710 Serious about music .666 Willing effort .637 .360 Feel emotion* .439 .382** Symbols and markings .813 Counting music .767 Reading music .755 Notes and rhythms .746 Sense of rhythm .554 Steady beat .384 .462 .347 Love listening .715 Music is fun .704 137 Table Appendix F (continued) Variable Effort Ability Affect Class Background Comp 6 Comp 7 Caring about music* 321** .657 Like to make music .647 Please others .509 Naturally Creative .455 Teacher favoritism .737 Teacher temperament .729 Teacher understands you .654 Liking teacher .608 .335 Liking other students .568 .448 Getting along with others .547 Musical relatives .783 Runs in family .765 Musical parents .719 Starting young .458 .425 Being with friends* .330* .428 Natural talent .762* Natural ability .737* Liking sound* .432** .616 Good ear* .347 .538 138 Table Appendix F (continued) * = variables that did not load onto the expected factor ** = secondary factor loadings > .3 that are consistent with the Asmus model 139 APPENDIX G Rotated Component Matrix for 34 AMF Variables Suppressed to Five Factors Variable Ability Effort Background Affect Class Counting music .754 Symbols and markings .753 Notes and Rhythms .745 Reading music .711 Sense of rhythm .643 Steady beat .608 Good ear .470 Goals and practice .734 Practice effort .730 Trying hard .714 Practicing .705 Serious about music .655 Willing effort .652 Runs in family .785 Musical relatives .715 140 Table Appendix G (continued) Variable Ability Effort Background Affect Class Musical parents .687 Starting young .609 Natural talent .596 Natural ability .587 Being with friends .429 .407* Music is fun .720 Love listening .703 Like to make music .644 Caring about music .358* .624 Liking sound .510 Feel emotion .479 Please others .450 Naturally creative .430 Teacher favoritism .675 Liking teacher .673 Teacher temperament .664 Liking other students .660 Teacher understands you .647 Getting along with others .621 141 Table Appendix G (continued) * = secondary factor loadings > .3 that are consistent with the Asmus model