Florida State University Libraries

Transcription

1 Florida State University Libraries Electronic Theses, Treatises and Dissertations The Graduate School 2006 The Effect of Performance Medium on the Emotional Response of the Listener as Measured by the Continuous Response Digital Interface David Scot Plack Follow this and additional works at the FSU Digital Library. For more information, please contact

2 THE FLORIDA STATE UNIVERSITY COLLEGE OF MUSIC THE EFFECT OF PERFORMANCE MEDIUM ON THE EMOTIONAL RESPONSE OF THE LISTENER AS MEASURED BY THE CONTINUOUS RESPONSE DIGITAL INTERFACE By DAVID SCOT PLACK A Dissertation submitted to the College of Music in partial fulfillment of the requirements for the degree of Doctor of Philosophy Degree Awarded: Summer Semester, 2006

3 The members of the Committee approve the dissertation of David Scot Plack defended on April 26, Clifford K. Madsen Professor Directing Dissertation Eric Ohlsson Outside Committee Member John Geringer Committee Member Diane Gregory Committee Member Patrick Dunnigan Committee Member Jayne Standley Committee Member Approved: Don Gibson, Dean, College of Music The Office of Graduate Studies has verified and approved the above named committee members. iii

4 ACKNOWLEDGEMENTS Dr. Eric Ohlsson, and Dr. Jayne Standley: It was an honor to have you serve on my committee. You both are among the scholars and musicians whom I have looked up to and admired for many years. Your guidance, expertise, and support were essential to this process. Dr. Sean Murray: I cannot thank you enough for helping me in the editing of the renditions, something that became a major obstacle in the study. Your generosity in helping me did not go unnoticed or unappreciated. Thank you, my friend. Professor Diane Gregory: Thank you for agreeing to serve on my committee. Further, I want to thank you for your time helping me with the preparation of the CRDI stations and study room, and for your extensive knowledge in the use of the CRDI and in this area of study. Dr. John Geringer: I simply cannot imagine pursuing this degree at a better time, coinciding with your return to FSU as an eminent researcher and scholar. Your classes were instrumental to my doctoral education and I cannot imagine a better or more enjoyable person to learn from. For all of the help and guidance you offered, thank you. Dr. Patrick Dunnigan: You have looked out for me and steered me in the right directions since the first day we met, years and years ago. You cannot imagine the feelings of gratitude, admiration, and appreciation I have for you. You are a true mentor and a true friend. Thank you for always looking out for me. Dr. Clifford Madsen: At the end of this process, I have had time to sit back and think how fortunate I am to be one of the select students in higher education, anywhere, to have had the privilege of studying under your direction. Each day, the scope and perspective of this privilege dawns on me and makes me more and more thankful for this opportunity. Thank you for all that you have done for me over the years. You have truly taught me life lessons that I will never forget. iv

5 TABLE OF CONTENTS List of Tables... vi List of Figures... vii Abstract... viii 1. INTRODUCTION REVIEW OF LITERATURE... 3 Music Listening... 3 Music Preference... 4 Select Populations, Musical Training, and Musical Preferences... 5 Race, Culture and Musical Preferences... 7 Musical Elements and Musical Preference... 9 Non-Musical Elements and Musical Preference Exposure, Familiarity, Repetition and Musical Preferences Genre, Style and Musical Preferences Music Tension Studies Affective, Aesthetic, and Emotional Response Studies Measurement of Affective, Aesthetic, and Emotional Response Population Research and Affective, Aesthetic, and Emotional Response Musical Elements and Affective, Aesthetic, and Emotional Response Non-Musical Elements and Affective, Aesthetic, and Emotional Response Musical Genre and/or Specific Compositions and Affective, Aesthetic, and Emotional Response Performance Media and Affective, Aesthetic, and Emotional Response Statement of the Problem METHOD RESULTS v

6 5. DISCUSSION APPENDICES A. Composite Response Graphs B. Consent, Instructions, and Questionnaire REFERENCES BIOGRAPHICAL SKETCH vi

7 LIST OF TABLES Table 1: Release Information for Stimulus Recordings Table 2: Order of Excerpts vii

8 LIST OF FIGURES Figure 1: Standard Continuous Response Digital Interface (CRDI) Dial Figure 2: Composite Response Graph - All Excerpts Figure 3: Composite Response Graph - Voice Excerpt Figure 4: Composite Response Graph - DCI/Marching Band Excerpt Figure 5: Composite Response Graph - Wind Ensemble Excerpt Figure 6: Composite Response Graph - Popular dance Music Excerpt Figure 7: Composite Response Graph - Piano Excerpt Figure 8: Composite Response Graph - Composite Overall Mean and Standard Deviation Response Graph viii

9 ABSTRACT The primary purpose of this study was to determine whether participants from varied performance-media backgrounds experience a felt emotional response through performance media different from their own. Specifically, how does this emotional response compare to the emotional response shown through their own performance medium? Secondarily, the study examined whether the response patterns over time of the Continuous Response Digital Interface (CRDI) dial show a relationship between listening groups and/or performance media. Volunteer participants (N = 143) consisted of graduate and undergraduate music majors, non-music majors, and non-musicians at a large comprehensive university. Recordings were presented using the following performance media: (1) voice, (2) wind ensemble (non-marching), (3) marching band (non-music major), (4) piano, and (5) popular dance music (non-musicians). Based on primary performance area, participants were assigned to one of the five groups: voice (n = 31), wind ensemble (n = 25), marching band (n = 27), piano (n = 33), and non-musicians (n = 27). Participants in each group were asked to manipulate the CRDI dial corresponding to their felt emotional response to the music. All participants listened and responded to five performance media renditions of Giacomo Puccini s Nessun Dorma from the opera, Turandot. Means and standard deviation were determined and graphically displayed. Graphs include separate composite means and ongoing standard deviation response graphs for each excerpt and one composite response graph detailing overall mean and overall mean standard deviation by group for each excerpt. Visual inspections of the composite graphs demonstrate both large as well as subtle differences among and between groups and that participant responses differentiated across the musical stimuli. Ongoing changing responses of participants within participant groups provide the most important description of responses. Further analysis of the composite response graphs demonstrates a strong relationship between responses of the music major groups (voice, piano, and wind ensemble) and their respective corresponding excerpts. There was also a strong relationship between the responses of the non-music major groups and their respective corresponding excerpts (DCI/marching band and popular dance music). ix

10 CHAPTER 1 INTRODUCTION For decades, research focusing on emotion in music and music preference has been examined and explored by music researchers. These broad areas of study in music can provide the scholar with both dilemmas and opportunities in the quest to find answers to some of the most common and basic questions about emotion in music and musical preference. What does one person hear and feel listening to music that another simply does not? How is emotion in music heard and felt from one individual to the next? What determines an individual s musical preference? Can these preferences be altered? How often does the individual s preference change, if at all? Answers to these questions, while general in nature, are some of the most sought after answers in music research. The present study will further investigate the areas of emotion in music within the context of performance medium. Rapid changes in technology are providing increasing opportunities for both music consumer and music researcher. For the consumer of music, changes and advancements happen so rapidly that it is difficult to keep up-to-date with the ever-changing consumer market. In addition, music today is more readily sought after than ever before through the Internet and through other affordable and accessible means. Where once consumers of music were limited in the style, genres, and performance mediums easily available to them, they now find a virtual smorgasbord of music perhaps once unfamiliar and unattainable to them. For evidence of this, one needs to look no further than the growing popularity of world music and other genres now gaining more attention. Further, personal music devices are now portable with the invention and advancement of MP3 players, the Apple ipod, and other personal listening devices. Indeed, music is ubiquitous and appears in different genres and styles than previously available. Along with this wider variety of genres and styles available, have the possibilities for emotional and aesthetic experiences notably increased? Or, are these opportunities for emotional and aesthetic experiences still only limited to what is familiar and comfortable to the listener? This study was designed to determine if participants from varied performance-media backgrounds have what they consider a felt emotional response while listening to several renditions of Puccini s Nessun Dorma through performance media the same as and different 1

11 from their own and to examine the emotional response of each of five participant groups. Further, the study examined whether the response patterns over time of the Continuous Response Digital Interface (CRDI) dial show a relationship between listening groups and/or performance media. 2

12 CHAPTER TWO REVIEW OF LITERATURE In an effort to concentrate research literature towards the focus of this study, the following review of literature will cite studies of varying nature and design from different areas of research including: (1) music listening, (2) music preference, (3) perception of tension in music, and (4) affective, aesthetic, and emotional response in music. Music Listening Music listening seems vital to understanding music and the importance and meaning it holds to the individual listener. For one to explore more advanced areas of research, one must first examine how people listen to music. In 1999, Sloboda, through a survey regarding listening habits, investigated the everyday uses of music and noted that the listener was an active agent who chose music according to their specific needs and purposes. He also noted that music selected by the individual is often used to enhance solitary activities and that listeners were generally ambivalent to music played in public areas where they did not have a choice as to what music was being played. While Sloboda s 1999 survey was generally focused on music listening, other researchers have more specifically investigated this area. Boal-Palheiros and Hargreaves (2001) investigated the differences between the functions of music listening at home and at school and the potential effects of age and nationality on these differences. Their findings showed that music listening was an important leisure activity and that most children showed moderate positive attitudes towards school music. Simply stated, home listening was linked with enjoyment, mood, and social relationships while school listening was associated with motivation for learning. These results are of importance to the school music teacher as they develop and plan a course for active listening participation in their music classrooms. Hedden (1980) assists school music teachers by presenting a plan and course of action for developing music listening skills. In his plan, Hedden listed notated themes or visual representations, guided listening, and other methods or teaching techniques designed to develop the aural acuity and motivation of the listener. This is not a new concept as many elementary and secondary-school teachers employ techniques such as these designed to motivate the listener and 3

13 guide them to a better appreciation and understanding of music. There have been several studies that focus on music listening and school-aged children. Focusing on the development of listening habits, in an extensive longitudinal study of music listening of German children aged eleven to seventeen, Behne (1997) examined whether individual biographical events during these years affected listening habits and preferences. He found that the development of musikerleben or musical life had to do with coping during these adolescent years of life. Stratton and Zalanowski (2003) studied the daily music listening habits of college aged students by having them keep a music-listening journal. Surprisingly, their study showed that nonmusic majors listened to music more the music majors, that the preferred style of music was rock, and that most listening occurred around day to-day events such as studying, dressing, and driving. While some studies have focused on the development and habits of individual music listening, other research focuses upon listeners perceptions in music. Music Preference Studies Research in music preference has continued for many years. Why we listen to music and how we formulate our preferences are often the subject of this research. However, even with the intense focus of research in this area there still remain unanswered questions in the study of music preferences (LeBlanc, 1991). Through a series of six studies examining beliefs about music, Rentfrow and Gosling (2003) found that people consider music an important aspect of their lives and that they believed listening to music was an important activity they engaged in frequently. Comparing 3,500 individual preference responses to a wide array of self-view personality dimensions, researchers found four music preference dimensions: Reflective and Complex, Intense and Rebellious, Upbeat and Conventional, and Energetic and Rhythmic. One of the major questions facing researchers is whether individual music preferences can be altered or modified. In 1989, Finnas provided a thorough review of music preference research conducted up to that point in time. The review focuses on factors influencing musical preferences of young people who had received general musical training. These factors included specific characteristics of music, the effect of familiarity with music, and the effect of social influence. Hargreaves, Messerschmidt, and Rubert (1980) examined these factors in a study of musical preference through both affective and evaluative ratings. On both scales, musically trained subjects gave significantly higher ratings than untrained subjects and familiar pieces received significantly 4

14 higher ratings than unfamiliar pieces. For ratings of quality, classical musical selections received significantly higher ratings over popular music selections. However, liking responses favored popular pieces over classical musical selections. Finally, there was a high correlation between individual ratings of quality and liking; allowing researchers to generalize that higher preferences existed for higher quality music. As the study of musical preferences has developed the question of measurement has been raised as a key issue. In 1982, Boyle examined the relationships among college students responses to three music appreciation tests: the subtest of Wing s Standardized Tests of Musical Intelligence, the musical sensitivity portion of Gordon s Musical Aptitude Profile, and the Indiana-Oregon Music Discrimination Test. Although these tests have been used in the area of musical preference, the need to develop more specific response measures is apparent. Prior to the development of the Continuous Response Digital Interface (CRDI), preference-based response measurements where primarily limited to conventional static measurements, such as the Likert scale. Brittin and Sheldon (1995) compared listeners continuous (CRDI) ratings and static (Likert) ratings of preference of art music. In addition to the comparison of response measurements, researchers also examined style and performance media through this study. Participants included 100 music majors and 100 nonmusic majors who were asked to rate 12 excerpts of Western art music. The musical stimuli included excerpts in three substyles (Baroque, Romantic, and 20 th century), which were performed through three different performance media (strings, keyboard, and winds). Results showed that music majors preference ratings were significantly higher than the preference ratings of nonmusic majors. Further, results did not show a significant effect between the different uses of measurement, however; there was a significant interaction between type of measurement and major. Music majors static and continuous ratings of preference were similar, while nonmusic majors continuous ratings were much higher than were their static ratings. Select Populations, Musical Training, and Musical Preferences Many researchers have been interested in the study of musical preferences for different aged populations. Smith (1994) analyzed musical preferences according to the decade of birth from the 1920 s through the 1970 s. Analysis showed that generations tend to assume the music style of their youth and that they tend to maintain this preference throughout adulthood. 5

15 Although there a few genres that maintain generational popularity, most genres are most preferred by those who were young at the time of that genre s peak of popularity. It appears that this generalization cannot be understated when discussing the importance and relevance of research dealing with the musical preferences of young people and other populations of listeners. LeBlanc, Sims, Siivola, and Obert (1996) studied the musical preference ratings of 2,262 listeners aged 6 91 years, grades 1 through college, in addition to adults who were not college students. Musical stimuli in this study included 18 examples of art music, traditional jazz, and rock, which were played in a random sequence. Listeners were asked to rate each listening example immediately after hearing it. Results showed that preference mean ratings were comparatively similar across all grade levels for each style. In addition, style subtests were pooled to make a general index of music preference in which a characteristic pattern of response emerged across grade levels. Listeners in grade one had a high level of preference but levels declined to a low point in grade six, after which preference steadily rose into the high school years where it reached its highest point at the college level. In the adult group, preference declined again, however, adult preference was higher than that of any other grade level except grade one and the college level. Other research examined the musical preferences of younger listeners. Finnas (1989a) compared young listeners privately and publicly expressed musical preferences, Siebenaler (1999) examined elementary students song preference, including familiarity, for several songs published by the Music Educators National Conference, and LeBlanc (1979) developed a listening test to measure generic style music preferences of fifth-grade students. While research regarding age and musical preferences is important to the understanding of our musical preferences, one must also look at musical training to gain an additional perspective. Geringer (1982) focused on the relationship between age and musical training on verbal and operant music listening preferences. Results showed that although responses regarding preferences among music major participants were consistent, verbal and operant responses for both elementary and undergraduate nonmusic majors were not. Results in a study examining the effects of age, gender, and musical training on musical preferences of British secondary school students showed that most students, as they grew older, did not prefer classical music (Hargreaves, Comber, & Colley, 1995). However, results also indicated that females displayed a greater acceptance for various musical styles than males in this age group. 6

16 While there is research examining preferences ratings of elementary age students, other research has focused on the musical preferences of older listeners. In a study that identified the narrow musical options provided to 242 members of both select and non-select high school choruses, Rentz (1994) found that both groups preferred popular music over other genres. Reynolds (2000) examined the relationships between musical aptitude and musical preference among high school students. For the study, Reynolds developed an aptitude profile and preference test to administer to 204 high school participants. Results showed that students with different musical preferences do not significantly differ in tonal or rhythmic aptitude, reinforcing LeBlanc s preference model, which excludes aptitude as a variable affecting preference. In a study focusing on older participants, Lychner (2000) compared the preference of baby boomers to the preferences of undergraduate non-musicians. The undergraduate population was selected such that no participant was a music major or professional musician while the baby boomer population was selected such that no participant was a professional musician. Both groups, selected haphazardly, contained an equal number of males and females. Participants were asked to complete a survey regarding their musical preferences. Results indicated that different genres of music were equally preferred and that there was a significant relationship between gender and genre choice. Preference of nearly half of the total number of respondents fell into either the eclectic or rock genre categories. Males most often preferred rock music while females preferred pop or were more eclectic in their musical preferences. Race, Culture and Musical Preferences As school environments are increasingly more diverse and multi-cultural, the need to research differences in musical preferences relating to race, ethnicity, and culture is apparent. McCrary (1993) examined the effects of black and white listeners and performers race on musical preferences. Listeners included middle school-aged children and university college students. Only black listeners showed significantly different ratings in their preferences for black and white performers. These listeners also rated black performers highest when they identified the performers as being black. Ratings of the white listeners were virtually equal for the black and white performers. Following with a similar study, Morrison (1998) added the element of visual/aural condition to the study, allowing students to visually identify the race of the performer whose music they were listening to and rating. Results in this study indicated that 7

17 white listeners preferred music performed by white artists regardless of presentation condition while black listeners preferred examples by white performers when presented in the music-only condition. In a study examining the effect of age, country, and gender on musical preferences, results showed that each variable showed significant influences on preferences of school-aged students in Greece, Korea, and the United States (LeBlanc, Jin, Stamou, & McCrary, 1999). However, as there were significant two and three way interactions between variables, researchers concluded that generalizing findings in musical preferences from one culture to another must be exercised with caution. It is not uncommon for musical artists to find cross-cultural success. For instance, Western artists are very popular within the Japanese market, especially with Japanese adolescents. One may find more and more of a Western influence in Japanese popular music. Hiroshi and Wells (1998) studied the genre preferences of Japanese adolescents for Western popular music. Results indicated, the musical preferences of Japanese youth are neither a matter of monolithic mass tastes or of idiosyncratic individual choice (p. 41). They suggest that the Japanese youth embrace numerous genres in such a way that their musical preferences may seem indistinguishable. Darrow, Haack, and Kuribayashi (1987) similarly examined descriptors and preferences for Eastern and Western music by Japanese and American students. Participants, 487 nonmusic majors from the United State and Japan, were asked to listen excerpts of both Western and Eastern music styles. Following the listening, participants were asked to describe the character of the selection by selecting one of nine adjectives. In addition, listeners were asked to rate their preference on a 7-point Likert scale. Results determined that both groups indicated an overall preference for Western music; however, Japanese students were more open to Eastern music styles than American students. Further, students with greater musical training generally used a wider range of preference than those with little to no musical training. Similar results were found with students from the United States, Hong Kong, and the People s Republic of China (Morrison & Cheung, 1999). In a study combining the investigation of cultural music preferences and response modes, Brittin (1996) analyzed the preference responses of 75 music majors, 75 nonmusic majors, and 75 junior high school musicians. Participants listened to excerpts representing various cultures, including the Caribbean, Africa, India, and Oceania. Preference responses were recorded using a CRDI, representing continuous measurement, and through 10-point Likert scales, representing 8

18 static response measurements. Listeners using the CRDI rated excerpts significantly higher than those recording via the Likert scale. In addition, no significant difference in preference was recorded in relation to musical experience and expertise. Musical Elements and Musical Preferences A large amount of research in the area of music preference isolates elements of music to examine their effect on individual preferences. Many researchers have used tempo as the independent variable in their studies. Other studies examine individual preferences for a specific musical element. One such study examined the tempo preferences of different age music listeners (LeBlanc, Colman, Mccrary, Sherrill, & Malin, 1988). In this study, researchers measured the effect of four levels of tempo on the self-reported preferences of 6 different age groups, from third grade through college level, for traditional jazz music listening examples. Analysis of variance results showed a significant preference for increasingly faster tempi at every age. Age showed a strong influence on overall preference ratings. Highest ratings were among the youngest listeners in the third grade and the ratings steadily declined to a low point with the seventh grade listeners, rising again as age increased up to the college level. LeBlanc and McCrary (1983) and Montgomery (1996) each examined the effect of tempo on music preferences of younger aged students. Results in both studies showed a strong and significant relationship between faster tempi and preference. Examining the cross-cultural boundaries, LeBlanc et al. (2000) tested the music preferences of younger aged listeners in Brazil, China, Italy, South Africa, and the United States. Once again utilizing jazz music gradually increasing in tempo, the study showed a highly significant relationship between faster tempi and higher preference ratings across all age groups. In addition, the effect of country and age were also significant influences on preference ratings. Brittin (2000) examined the preference and perception of tempo on rating scales with participants ranging from the second grade through sixth grade. Participants listened to children s songs sequenced with electronic keyboard accompaniments in style ranging from popular to more traditional folk styles. Three different versions of the survey combined different styles and tempos with each selected song. Consistent with previous findings, results showed a significant correlation between faster tempi and higher preference ratings across all grade levels except with second grade listeners. Students also preferred styles labeled Hip-Hop, Heavy Rock 9

19 Shuffle, Samba, and Funk while lower preference ratings were shown for styles labeled Polka, March, Bluegrass, and Piano Chords. Some research has combined the effect of tempo and tempo preferences with other musical elements such as pitch to analyze possible relationships. In one such study investigating pitch and tempo preferences, Geringer and Madsen (1987a) researched these preferences within the context of popular music. Participants listened to and rated original excerpts of popular music and excerpts altered in pitch and in tempo. In most cases, participants preferred the original, unaltered versions of each popular music selection to versions altered in pitch and tempo. These results are in contrast to the above findings showing that increased tempo relates to increased preference ratings. Flowers (1988) examined the effects of tempo, teaching and learning experiences, and mode on undergraduates and children s symphonic music preferences. Consistent with research in this area, both children and undergraduates gave higher preference ratings to excerpts with fast tempi than to slow tempi. Wapnick (1980) examined the variables of tempo, pitch, and timbre in preferences for recorded piano music. Each variable was assessed independently from the other two within the listening situation. 96 participants, randomly assigned to four groups, heard 12 excerpts of familiar and unfamiliar solo piano music. Results showed listener preferences for fast tempo and bright timbre when each variable was examined independently. Another musical variable and isolated element seen in the research of musical preferences is dynamics, because contrasts in dynamics could have a direct effect on the preference ratings of the listener. One such study focused on the expressive variations in dynamics and the relationship to musical preferences in elementary-aged listeners (Burnsed, 1998). Listeners heard and rated two versions each of 10 folk songs; one version of each song used expressive dynamic variation while the other was consistently even in dynamics. Results indicated a significant preference for the expressive variations of each song. In an earlier study with middle school students, Burnsed and Sochinski (1995) found similar results as a significant portion of the middle school students preferred the versions containing expressive dynamic variations over those without variation. In what started as a replication of the previous two studies, Burnsed (2001) provided excerpts which were considered more realistic in their representation of expressive dynamic contrasts to elementary-aged students, middle school students, and conductors. Results showed that age and musical experience might affect perception and 10

20 preference for subtle dynamic nuances in music. It did seem, however, that elementary-aged students might not have been able to detect the subtler dynamic nuances than in previous studies where more explicit contrasts were employed. This prompted the researcher to implement the new study with older students. Non-Musical Elements and Musical Preference While numerous studies cited use specific musical elements as an independent variable, it is not uncommon for non-musical variables to be studied and examined as it pertains to the study of musical preference. These unique variables often vary from study to study. North and Hargreaves (1997) investigated the effect of physical attractiveness on responses to pop music performers and their music and found that preference for popular music and its accompanying aesthetic response can be influenced and affected by the performer and their perceived physical attractiveness. In another study, Killian (1990) compared the effect of modeling on the development of musical preference. Junior high school participants of varying ethnic diversity were asked their preferences for performers (21) in the video of We Are the World and asked which solo they would most like to sing themselves. Results showed that students generally preferred same-race and same-sex models. Another study examined the relationship between humor perceived in music and preferences of different aged listeners and found that higher levels of perceived humor was associated with higher levels of musical preference (LeBlanc, Sims, Malin, & Sherrill, 1992). Some studies have focused attention on the effect of peers, teachers, and others on the musical preference. Furman and Duke (1988) examined the effect of majority consensus on music and nonmusic majors preferences for recorded orchestral and popular music. Their results indicated that nonmusic majors preferences were significantly affected by the preference of others. Alpert (1982) investigated the effects of disc jockey, peer, and music teachers approval on attitudes and preferences of fifth grade students. Results showed that music teacher and disc jockey approval for classical music increased classical music preferences among younger participants. 11

21 Exposure, Familiarity, Repetition and Musical Preferences Three independent variables often focused upon in preference research are exposure, familiarity, and repetition. Does the listener prefer more familiar music? How are preferences changed with repetition? Hargreaves and Castell (1987) examined the effect of familiarity on liking for melodic sequences among participants in each of six age groups and found that preference ratings revealed an inverted-u relationship with increasing age for familiar melodies. Peery and Peery (1986) studied the effects of exposure to classical music on the musical preferences of preschool children. Participants preferences for six classical music pieces and two popular music pieces were pre-tested, in which all participants liked all of the selections. Over a ten-month period, participants in the experimental group received a weekly music appreciation class. The post-test results from this longitudinal study indicated that the group receiving music appreciation classes preferred the classical music selections while the control group who had received no music appreciation class work, actually showed a decline in preference ratings. In another similar study concentrating on effects of appreciation and repetition, 30 seventh grade classes were divided into three groups, which were pre- and post-tested (Bradley, 1972). The experimental group received music appreciation class work, the second group was exposed to repetitive listening, and the third operated as the control group. Consistent with research in this area of study, results showed a significant change in music preference ratings for contemporary music in the first and second groups. Bradley preceded this study with one yielding similar results (1971). Price (1998) examined the effect of a music appreciation course on preferences of college nonmusic majors favorite composers. Participants were asked to rank their ten favorite composers before and after a 10-week course in music appreciation. At the end of the study, participants showed a greater tendency to list composers of the formal tradition, which was the focus of the appreciation course. One alternative explanation of these results is that this change in composer rating was a result of familiarity and not a greater liking for the composer s music. Bartlett (1973) randomly divided 149 college students, nonmusic majors, into three groups to examine the effect of repetition on the preferences and affective response for classical music. Each group was pre- and post-tested for liking of the musical selections. Results showed that repeated listening diminished liking for the best-liked selections; however, repeated listening increased liking for the classical selections. 12

22 Two studies examined the effects and combination of familiarity and exposure with musical complexity on musical preference ratings (Heyduk, 1975; Hargreaves, 1984). Both studies utilized predictions in complexity derived from an inverted-u theory of the relationship between complexity and liking for musical selections. Heyduk examined the preference ratings of 120 undergraduate participants as they related to musical complexity and exposure frequency. Listeners heard four piano compositions constructed to represent differing degrees of complexity. Results indicated that listeners liking increased with musical complexity, consistent with previous findings in this area. Further, one of the four selections was selected and rated another 16 times with results varying depending on the listener s preferred complexity level. Hargreaves (1984) focused on the effects of repeated listening on the liking of pieces of different styles and varying complexity. Using easy listening, popular, classical, and avant-garde jazz selections, each designed to gradually increase in subjective complexity, results indicated that changes in preference and liking that occurred as a result of repetition supported the inverted-u complexity hypothesis. Burke and Gridley (1990) investigated the preferences of college students, some with extensive musical training and some with fewer than two years total. Listeners heard recordings of Bach, Debussy, Grieg, and Boulez, which were previously rated and ranked for subjective complexity by seven music professors. Once more, when preference ratings were compared to ratings of complexity, an inverted-u curve was obtained. Furthermore, liking was higher among students with greater musical training and backgrounds. Genre, Style, and Musical Preferences Through examining the relationship between genre preference and listener type, Stein (1985) stated that the experience of music is shaped by the music s genre and that music expresses information about personality (p. 98). Every individual develops, through musical experience and life, genre and style specific preferences in music. Are genre preferences based on musical experiences over our lifetimes? Jin (1999) discusses these ideas among others while investigating the relationship between listeners music experience and their preferences for specific genres and styles of music. In addition, Jin discusses the relation between familiarities with a certain style of music, the effect of the listeners valuation of the music genre, and the effect of parents, age, gender, and peers on our development of preference for genres and styles of music. 13

23 One study examined the idea of musical experiences and its relationship with genre preferences. Pickert (2002) studied participants aged who were actively involved in a genre-specific music ensemble. Participants, all German, were active members in chamber music groups, rock bands, orchestras, jazz bands, or in other smaller ensemble settings. Results showed a high correlation between active or passive contact with music and preference ratings. Individuals seem to listen to genres and styles more familiar to them through different levels of active and passive participation in music; members of rock bands tend to prefer the rock genre and so forth. In an earlier similar study, Pickert (1993) surveyed 93 music teachers of state schools in Germany about what genres of music they participate in, in ensembles outside of the classroom. In comparison to their own personal likes and dislikes in regard to music genres, these teachers yielded a narrow range of music preferences, with a significant correlation between preferences and outside the classroom participation in music. The role and status of popular music in society is also important. The business side of popular music is a multi-billion dollar industry so it should not be surprising that what people like to listen to, and why they like to listen to it, would be a focus of attention for researchers. In one study, self-reported preference responses of students in grades five through college were compared (Boyle, Hosterman, & Ramsey, 1981). Participants, many with different backgrounds and experience, were asked reasons for their pop music preferences. Elements of pop music most commonly referenced as important were melody, mood, rhythm, and lyrics. Pop music is not the only genre focused on by researchers interested in determinant factors of music preference. Orr and Ohlsson (2001) tested the correlation between preference and complexity in jazz and bluegrass. Participants judged both the complexity and liking for excerpts performed by musicians instructed to improvise at different levels of complexity. Results were consistent with previous research supporting the inverted-u shaped relationship between liking and complexity in the bluegrass genre, but not for the jazz genre. In research investigating the effect of overt categorization on preference for popular music styles, Brittin (1991) found that subjects with musical experience rated popular music styles including pop, rock, jazz, and crossover excerpts higher than listeners without musical experience, while the preference ratings of females were significantly higher than were male listeners. Musical experience and gender were found to significantly affect musical preferences for the genres of pop and jazz, but not for the rock genre. 14

24 Music Tension Studies Some musicians believe that true tension and release in music is a predominant characteristic that affects the individual such that their preferences and emotions are formed and established. Madsen and Fredrickson (1993) replicated Nielsen s research (1983) on musical tension in an attempt to determine if the idea of musical tension was universal and to determine if tension in music can be measured through varying instrumentation. Madsen and Fredrickson used the Continuous Response Digital Interface (CRDI) to replicate Nielsen s earlier study. Results supported the notion that Nielsen s findings were replicable through the use of different measurement techniques, including the CRDI. In a study focusing on the perception of tension in music, Fredrickson (1997a) asked elementary, middle, and high school students to listen to a recording of Haydn s Symphony No. 104, first movement. He measured their responses to tension with the CRDI dial. Without providing a definition of the term tension, students were asked to move the dial corresponding to their perceptions of tension. Results showed that while listening, the younger participants employed a wider range of the interface dial than did the older participants. The generalization of the results from this study reinforces the notion that tension may be defined somewhat differently across all population groups. Fredrickson continued research in this area with a study designed to measure responses of older-aged adult musicians and non-musicians (Fredrickson, 2000). Similarly to the aforementioned study, Fredrickson asked participants to listen to different musical stimuli and indicate their perception of musical tension through the dial on the CRDI without being provided a standard definition for what they were listening for. Results indicated that musicians and non-musicians responded similarly when tension and release were at its peak within a musical context. While some studies focus on different populations and their perceptions of tension in music, others examine the effect specific musical elements may have in the perception of tension. Fredrickson (1994) explored the effect of a conductor on the perception of tension by providing participants the ability to listen to a piece of music with and without the visual stimulus of a conductor. Although findings were not statistically significant, results did show those who watched the conductor had slightly different perceptions than the aural only group. In a study on the effect of performer use of rubato on the perception of tension, Fredrickson and Johnson (1996) found that no significant differences in the perception of tension were recorded 15

25 by participants listening to two pieces, one with and one without the use of rubato. Misenhelter (2001) isolated dynamics in Beethoven s Symphony No. 7 in his study of tension in music. Music majors responded via the CRDI to dynamics while nonmusic majors responded to perceived tension via the CRDI. As both groups responded similarly, findings suggest that contrasts in dynamics may have a direct effect on the perception of tension in music. In a similar study investigating the effect of dynamics on perceived tension, Burnsed and Sochinski (2001) asked participants to listen to Haydn s Symphony No. 104 and record their responses on a newly designed Tensiometer. Results were similar to other studies involving dynamics and the perception of tension as responses indicated a strong relationship between score marked dynamics and tension responses. While some research has focused on musical elements such as the conductor, rubato, and dynamics, other researchers have been interested in the perception of tension within a specific selection of music. In a study comparing responses from college student nonmusic majors to a study previously done with music majors, Fredrickson (1997b) used Percy Grainger s Irish Tune from County Derry as the musical stimulus. Nonmusic major CRDI responses compared similarly to the music major responses from the previous study leading to the generalization that while musical experience may help in the development of advanced sensitivity to different elements of music, it might not be necessary for a differentiated response. Fredrickson (1999) followed this study with a study utilizing Gustav Holst s First Suite in E-flat as the musical stimulus. The primary focus of the study was to determine if student musicians who both rehearse and perform a musical selection perceive musical tension differently than listeners who do not experience this process. Each participant listened to the recording of the Suite and registered their perceptions of musical tension via the CRDI before rehearsals began and following the final performance. Results indicated that the musical process of rehearsal and performance did not seem to affect the perception of musical tension. While most research tends to focus on classical and more traditional genres, Fredrickson and Coggiola (2003) examined musical tension responses of music majors and nonmusic majors to two selections of jazz. Using a CRDI, listeners responded to recordings of St. Louis Blues by W. C. Handy, one version sung by Nat King Cole and the other by Ella Fitzgerald. Consistent with previous research, the responses of the music majors and nonmusic majors did not differ significantly from one another but did show patterns unique to each rendition of the selection. 16

26 How does musical tension relate to emotional or aesthetic response? How similar are these ratings when compared with one another? Using the CRDI, Fredrickson (1995) measured musicians and non-musicians responses to both musical tension and aesthetic response in the first movement of Haydn s Symphony No Results demonstrated that musical tension responses were more variable than aesthetic responses. In addition, aesthetic responses were consistently higher and more positive than tension responses. Finally, non-musicians responses were consistently more positive than musicians responses. Madsen (1998) compared two responses simultaneously while listening to the first movement of Haydn s Symphony No The participants, 50 adult musicians, recorded their responses of arousal and affect using a twodimensional CRDI with the vertical dimension (arousal) reading Exciting and Relaxing and the horizontal dimension (affective) reading Ugly and Beautiful. Results of this study seemed to be consistent with previous findings. When analyzed within a musical context, arousal responses and affective responses seemed to correspond to variations in the music. In addition, results and graphic analysis demonstrated that these two areas shared an inverse relationship: the exciting-relaxing arousal dimension responses were often the opposite of the ugly-beautiful affective dimension responses. Affective, Aesthetic, and Emotional Response Studies Standifer (1970) states, Every normal human being has some measure of aesthetic sensitivity (p.112). The focus of research in this area centers on this idea, that aesthetic sensitivity can be monitored, measured, and developed in every individual. As research in aesthetics and affective response developed over the years, Price proposed a glossary of terminology (1986). Lychner (1998) followed the glossary with a study designed to investigate possible differences in the use of the terminology used when individuals were asked to respond to music. Using Puccini s Nessun Dorma as one music stimulus among other excerpts, results showed a strong similarity between the terms aesthetic response and felt emotional response while perceived tension was markedly different than the other conditions. As this area of research has grown in interest, there are many compilations and reviews of literature pertaining to the study of aesthetics and emotions in music that may provide insights (Swanwick, 1973; Sloboda, 1992; Sloboda, 1999; Juslin and Sloboda, 2001; Juslin and Zentner, 2002). 17

27 Bever (1988) provides a cognitive theory of emotion and aesthetics in music and argues that in the absence of a clear basis for arousal, listeners unconsciously select one of their own emotions as the cause (p. 165). In this way, listeners simply unlock emotions they already have when processing music as a complex object. In 1980, Russell provided a circumplex model of affect. Russell s model is made up of descriptive affective states: pleasure, excitement, arousal, distress, displeasure, depression, sleepiness, and relaxation. As many see these dimensions existing independently from each other, research suggests these dimensions are highly interrelated. Russell goes on to compare the task of affective self-report to one labeling photographs of faces. Each photograph has unique qualities and properties; however, they also share commonalities that help us conceptualize and build emotional states. Measurement of Affective, Aesthetic, and Emotional Response One of the significant advances in the study of emotional response to music was the development of the Continuous Response Digital Interface, the CRDI. One of the earliest studies utilizing the CRDI measured aesthetic responses of music faculty and advanced graduate students. In the first of a series of studies, Madsen, Brittin, and Caparella-Sheldon (1993) had listeners continuously move the dial on the CRDI as it corresponded to their perceived aesthetic response to Act I of Puccini s La Boheme. Participants also completed an exit questionnaire regarding their aesthetic experience. While there were different responses for each individual throughout the excerpt, there were heightened responses at peak moments in the excerpt. Participants indicated that the movement of the dial did correspond to their perceived aesthetic experience. Madsen (1997a) discusses issues pertinent to the measurement of emotional response in an overview of research involving use of the CRDI. Madsen points out that while more can be done in this area of study, the CRDI seems to be an accurate and valuable measurement tool. Gregory s (1995) study, summarizing and discussing reliability measures in research using the CRDI, supports the notion that the CRDI is an accurate and reliable measurement tool in the area of aesthetic and emotional response. As the reliability and functional status of the CRDI improved, a two-dimensional CRDI was developed. The two-dimensional CRDI, in comparison to the one-dimensional CRDI, allows researchers to monitor two dimensions of emotion simultaneously, such as happy/unhappy and excited/not excited (Tyler, 1996). Madsen (1997b) utilized the two-dimensional CRDI when recording responses of participants listening to Act I of 18

28 Puccini s La Boheme. Participants responded with two dimensions representing relaxing/exciting and ugly/beautiful. Analysis indicated that there were no differences to the responses in the ugly/beautiful dimension throughout the excerpt. Results indicated that one may need to be aroused in order to have an aesthetic experience. Population and Affective, Aesthetic, and Emotional Response Perhaps one of the more interesting areas of research is that of cross-cultural comparisons. Results of one such study, comparing listeners responses from European and Asian cultural backgrounds, indicated that affective responses are more culturally based in tradition rather than inherent in the qualities of the music (Gregory, A. H. & Varney, 1996). Goins (1999) provides general music teachers an overview of the influence of culture on emotional response to music. This is important, as the music teacher needs to take into account the cultural diversity of each class to provide a meaningful musical experience. Frega (2000) utilized the CRDI to examine and replicate cross-cultural aesthetic responses. Frega also discusses the importance of incorporating the CRDI and much needed research into countries where research and funding is scarce, but needs for music education changes are high. Madsen, Byrnes, Capperella-Sheldon, and Brittin (1993) were interested in comparing the emotional responses of non-musicians with the emotional responses from an earlier study of highly experienced and trained musicians. Participants listened to excerpts from Puccini s La Boheme, Strauss s Death and Transfiguration, Holst s First Suite in E-flat, Haydn s Symphony No. 104, and Mozart s Vesperae Solennes de Confessore in C Major. Listeners responded via the CRDI and also responded afterwards to a questionnaire asking whether or not each had an aesthetic experience and to what extent (magnitude) in comparison to other listening experiences. The overall responses of both groups were very similar, as results did not indicate a difference between musicians and non-musicians in frequency or magnitude of their aesthetic experiences. Musical Elements and Affective, Aesthetic, and Emotional Response In a study examining the effect of tempo and dynamics on the perception of emotion in music, Kamenetsky, Hill, and Trehub (1997) asked adult participants to listen to one of four possible MIDI versions of each of four musical excerpts. Versions of each excerpts included one 19

29 with no variations in dynamics and tempo, one with variations in tempo only, one with variations in dynamics only, and one with variations in both tempo and dynamics. Participants rated each version on a 7-point scale for likeability and emotional expressiveness. Results showed that variations in dynamics resulted in higher ratings on both measures while variations in tempo had no effect. Further, a high correlation was found between musical preferences and emotional expressiveness. Holbrook and Anand (1990) also investigated the effect of tempo and situational arousal on listeners affective response to music and found that an intermediate level of tempo lead to higher levels of affective responses. Sloboda (1991) analyzed the structure of music and its relationship to emotional response. Participants indicated that laughter, tears, shivers, and a lump in the throat were all felt at various times while listening to music. When respondents were asked where these emotions were located in a musical context, findings showed tears were likely evoked by passages with sequences and appoggiaturas and shivers were reliably caused by unexpected harmonies. Non-Musical Elements and Affective, Aesthetic, and Emotional Response The choice of music literature in classrooms, rehearsal halls, and on concert stages is an important decision for every music educator. Does the music chosen allow for a more attainable aesthetic experience for students? In a study investigating the effects of stage/jazz band versus concert band music literature on the development of musicality and aesthetic sensitivity, Anderson (1975) found that students exposed exclusively or partially to concert band literature were more likely to develop aesthetic sensitivity than stage/jazz band students according to the California Test of Aesthetic Judgments in Music. Does choral music or music with text have an advantage over music without words or text regarding the possibility of aesthetic experiences for the listener? Do words in a music context conveying emotion have effects upon the listener s emotional response? Gfeller, Asmus, and Eckert (1991) investigated the relationship between emotional response and music and text. Stimuli in this study included text alone, commercial-type background music, commercial-type background music with text, atonal music, and atonal music with text. Once assigned to one of the five experimental groups, participants responded via semantic differential scales and to the Asmus 9-Affective Dimensions. Results showed significant differences across the five experimental conditions for affective response and significant differences in mood from pre- to 20

30 post-testing. Similar results were found in another study investigating emotional response of trained musicians to verbal and music information (Gfeller & Coffman, 1991). Under the same conditions and measurements as the previous study, results indicated the pairing of text increased the perception of complexity, lowered preference ratings, and resulted in a decline in selfreported mood. Coffman, Gfeller, and Eckert (1995) found that when faced with music and music with text settings, females responded with higher complexity ratings and reported their feelings more intensely than males. Musical Genre and/or Specific Compositions and Affective, Aesthetic, and Emotional Response Using the two-dimensional CRDI to examine emotional content in choral music, Mulder (1997) found that listeners of different sexes, moods, levels of sophistication, and areas of specialization responded to music similarly. The two dimensions measured via the CRDI were happy/unhappy and excited/not excited. In addition, verbal reports provided by participants supported the data collected via the two-dimensional CRDI. The jazz music genre has gained recent attention by investigators regarding aesthetic and emotional response. Do listeners and participants of jazz music experience emotional responses similar to listeners and participants in other musical genres? Coggiola (1998) examined the effect of complexity in jazz music and jazz experience on the aesthetic response of musicians. Participants were assigned based on performance experience to either a jazz musician group or a non-jazz musician group. Listening to four jazz excerpts that varied in melodic complexity during the solo passages, participants manipulated a CRDI indicating their aesthetic response. Only for the most conceptually advanced excerpt was there a significant difference between experimental groups. Interestingly, the jazz musician group rated this same excerpt highest in their aesthetic responses. Goins (1998) investigated the effect of Pat Metheny s music on mood states of listeners. Participants in the first study listened to 8 excerpts and responded via two CRDI s; the left dial corresponded to eight mood adjectives while the right dial corresponded simply to aesthetic response. Participants in the second study responded via pencil and paper with the Hevner adjective circle. Results were similar from study one to study two despite the differences in measurement. With the exception of one excerpt, participants agreed on perceived 21

31 mood states. Yielding comparable results, Goins conducted a similar study using only the Hevner adjective circle as the measurement (Goins, 2000). Misenhelter and Lychner (1997) examined the responses of listeners while listening to a recording of either Satie s Trois gymnopedies or Chopin s Ballade, op. 47. Ninety-six participants manipulated a CRDI dial as it corresponded to their aesthetic experience. Results indicated that the Satie excerpt was generally flat in aesthetic response while responses to the Chopin excerpt were considerably more variable. As both pieces are stylistically different, it seems that responses might be related to different musical contrasts. Misenhelter (2001) also investigated undergraduate music majors and nonmusic majors emotional responses to excerpts of Stravinsky s Le Sacre du printemps (The Rite of Spring). Results indicated a significant difference between the responses of music majors and nonmusic majors. When analyzed within a musical context and consistent with research in this area of study, response graphs indicated patterns related to changes in musical textures. Performance Media and Affective, Aesthetic, and Emotional Response There has been research employing different applications of performance media and performance genre. Geringer (1992) analyzed dynamic contrasts across the performance mediums of choral, orchestral, and piano performances. While no significant difference in dynamic range was found between the three mediums, the study did show a greater dynamic range in crescendos than in decrescendos. The study did not yield a significant interaction between performance medium and direction of dynamic change. In his study of the effect of tempo, style, and performance medium on 5 th grade students music preferences, LeBlanc found a strong effect for style and for performance medium in the development of music preferences. Results also indicated a preference for faster tempos among younger participants (1981). LeBlanc and Cote (1983) furthered prior research by isolating the effects of tempo and performing medium on elementary-aged children s music preference. Consistent with previous research, they found that tempo was more of an influential determinant of musical preferences than performance medium. In addition to preferences for faster tempi, results also indicated preferences for the instrumental performance medium. While these studies focus more on younger listeners, it may be possible to transfer results to studies involving older participants. 22

32 Madsen (1997c) studied focus of attention to musical elements and aesthetic response. In an effort to determine which elements of music are most prominent in relation to aesthetic response, participants manipulated the CRDI dial indicating which of five musical elements commanded their attention while listening to Act I from Puccini s La Boheme. The five musical elements included melody, rhythm, timbre, dynamics, or a combination of the four. Results showed that those participants monitoring all five elements, most often registered movements of the CRDI for dynamics. Participants who were asked to only monitor one single element, registered the highest degree of response for melody, followed closely by dynamics. Finally, melody was the musical element most closely related to listeners aesthetic responsiveness. In a similar study, Madsen, Geringer, and Fredrickson (1997) used Haydn s Symphony No. 104 as the musical stimulus to investigate focus of attention and aesthetic response. In this study, results showed that rhythm was closely related to listeners aesthetic responsiveness. Gregory (1994b) used the Continuous Response Digital Interface (CRDI) to examine the music preferences for acoustic and synthesized transcriptions of middle-school students and college undergraduate music majors. Gregory alternated between traditional instrumentation and synthesized or acoustic transcriptions in the music of Bach, Mozart, and Copland. Results showed significantly higher preferences for traditional instrumentation over acoustic transcriptions in the music of Mozart and Copland. In his research investigating the effect of performance medium on the perception of musical tension, Fredrickson (2001) asked the following questions: When we listen to music does the medium have an impact on our perception? How does a piece of music change when we arrange it for a different medium? How does the change impact the listener? Fredrickson used Shostakovich s Festive Overture as the musical stimulus; the original orchestral version and a transcription for wind ensemble. Fredrickson found similarities in perception of tension between musicians and non-musicians and across the performance media of band and orchestra. However, non-musicians listening to the band performance exhibited a mean tension response indicating the possible influence of the performance medium at different times through the experience. He concluded that the study did indicate the presence of small-scale effects based in part, on performance medium. Gregory (1994a) analyzed listening preferences of undergraduate music majors, high school musicians, and sixth-grade students. Based on their performance background, listeners 23

33 were divided into one of four groups: (1) keyboards, (2) chorus, (3) band, and (4) orchestra. Participants heard excerpts presented in different genres and performance mediums including contemporary compositions, popular classics, elementary classics, and jazz crossover selections. Categories of musical stimuli included Hindemith, Stravinsky, Mozart, eclectic, Silver Burdett/Ginn Elementary Series, and crossover jazz. Each musical category had a representative keyboard, band, choral, and orchestra excerpt with the exception of crossover jazz, which were all instrumental in medium. Responses were recorded via the CRDI and showed instrumental biases in preferences among high school and college musicians when listening to relatively unfamiliar music. However, high school musicians tended to be more instrument-biased than college music majors, allowing the researcher to generalize that musical training may help lead to a more open view across music genres. Statement of the Problem This study was designed to determine if participants from varied performance-media backgrounds have what they consider a felt emotional response while listening to several renditions of Puccini s Nessun Dorma through performance media the same as and different from their own, and to examine the emotional response of each of five participant groups. Specifically, how does this emotional response compare to the emotional response shown through their own performance medium? Secondarily, the study examined whether the response patterns over time of the Continuous Response Digital Interface (CRDI) dial show a relationship between listening groups and/or performance media. 24

34 CHAPTER THREE METHOD There were two purposes to the present study. The primary purpose was to determine whether participants from varied performance-media backgrounds experience a felt emotional response through performance media different from their own. Specifically, how does this response compare to the response shown through their own performance medium? Secondarily, the study examined whether or not trends apparent through the manipulation of the CRDI dial show a relationship between listening groups and/or performance media. Participants Volunteer participants (N = 143) in this study consisted of graduate and undergraduate music majors, nonmusic majors, and non-musicians at a large comprehensive university. Volunteer participants for the study were recruited through sign-up sheets posted in the College of Music. Musician participants were those students pursuing a degree granted through the College of Music. Non-musician participants were those students pursuing any degree program offered by the university that was not granted through the College of Music and were students with no previous formal music training. Non-music majors were those students with formal music training, members of the University marching band, but were not pursuing a degree program offered by the University College of Music. Participants represented the following performance media: (1) voice, (2) wind ensemble (non-marching), (3) marching band (nonmusic major), (4) piano, and (5) popular dance music (non-musicians). Based on primary performance media listed above participants were assigned to one of the five groups. Participants in the voice group (n = 31) were music majors majoring in vocal performance or vocal music education. Participants in the wind ensemble group (n = 25) were instrumental music majors majoring in instrumental performance or instrumental music education and were not members of the University marching band, although they may have had some marching band experience in their past. Participants in the marching band group (n = 27) were non-music majors who were current members of the University marching band. Participants in the piano group (n = 33) were music majors majoring in piano performance, 25

35 accompanying, or were piano principals within the College of Music. Participants in the popular dance music group (n = 27) were all non-musicians with no prior formal music training. Conditions and Variables Participants in each group were asked to manipulate the CRDI dial corresponding to their emotional response to the music. All participants listened and responded to five excerpts derived from performance media renditions of Giacomo Puccini s Nessun Dorma from the Opera, Turandot. Each excerpt was edited and limited to two minutes in length. An effort was made to maintain basic musical form in terms of similar peaks and valleys. In addition, it was decided that each excerpt should be a complete entity in and of itself to preserve the composer and arranger s intention. To control for participant fatigue, the total stimulus provided to each participant was just over ten minutes in duration with 5 seconds of silence between each excerpt. Finally, in an attempt to control for possible order effect, the excerpts were presented in five counterbalanced orders. Each of the five orders was rotated for each listening group within the study regardless of the number of participants or group. One to eight participants participated in each listening group in the study. Musical Stimuli The musical stimulus used in the study was Giacomo Puccini s Nessun Dorma from the opera, Turandot. Nessun Dorma was selected because of its use in previous studies examining emotional and aesthetic response. Each performance media rendition was chosen because of its availability and differing musical qualities in each medium. From a total of ten renditions of Nessun Dorma performed through different and varying performance media, five were chosen in consultation with a panel of experts. Seven experts with an approximate total of 175 years of musical and research expertise participated in this selection process. The excerpts selected were as follows: The original voice rendition was that of Luciano Pavarotti and the London Philharmonic Orchestra, with Zubin Mehta conducting. The marching band/drum corps rendition was that of the Phantom Regiment Drum and Bugle Corps from their field show during the 1991 season. The wind ensemble rendition is a Patterson arrangement with the University of Houston Wind Ensemble performing, Eddie Green, conducting. Imperial Force 26

36 released the popular dance music rendition as a single only in Australia. John Bayless performed the solo piano rendition. Detailed information for the stimulus recordings is shown in Table 1. Table 1 Release Information for Stimulus Recordings Performance Medium Album Record Label Release # Voice The Essential Puccini (2004) Decca Records Drum Corps World Championships (1991) Drum Corps International 1991 Wind Ensemble the Meyerson (1996) Mark Custom Recordings UHWE-939 Popular Dance Single Release (1997) Odessa Mama Records OD0020 Piano (solo) The Puccini Album (1993) Angel Records Stimulus Recordings Once the different renditions of Nessun Dorma were selected, each excerpt was made to be two minutes in length with special consideration given to musical form and contour. In addition, each excerpt, with edits, represented a complete and separate musical entity with a clear beginning and introduction to a clear and compete ending. Input and output levels on all excerpts were normalized. All edits of recordings were done in the Sound Studio software program on an Apple Macintosh G4 Powerbook computer. When the edits were completed, each finished product was stored on an Apple Macintosh G5 desktop computer. To help account for possible order effects, excerpts were presented through five counterbalanced orders. In this process, excerpts were randomly assigned to one of the five order positions and were rotated through each of the five orders. In doing so, each excerpt was placed at a different position in each of the five orders as shown in Table 2, p

37 Table 2 Order of Excerpts Order A Order B Order C Order D Order E Voice Wind Ensemble Piano (solo) Popular Dance DCI/Marching DCI/Marching Piano (solo) Voice Wind Ensemble Popular Dance Wind Ensemble DCI/Marching Popular Dance Voice Piano (solo) Popular Dance Voice DCI/Marching Piano (solo) Wind Ensemble Piano (solo) Popular Dance Wind Ensemble DCI/Marching Voice Once the excerpts were edited and the counterbalanced orders set, five separate study CD s were made using an Apple G5 desktop computer. Each CD, with orders A through E, contained each of the five two-minute excerpts in one of the counterbalanced orders. Approximately five seconds of silence was placed between each excerpt. Listening Room A small soundproof room served as the listening room for this study. Curtains covered both a window and mirror and there were no other visual distractions on the walls. Lighting was provided through fluorescent ceiling lights. Eight CRDI stations were arranged on tables in the center of the room. Each station faced each other; they were visually separated from one another by partitions allowing each station to have privacy and respond without distraction. Each of the CRDI stations, numbered one through eight, included a CRDI dial, a set of stereo headphones, a pencil, and a questionnaire packet that included consent, instructions, a demographic questionnaire, and an exit question (see Appendix A). Sound reproduction included a Yamaha stereo amplifier and CD player. Data collection from the CRDI stations was recorded by an Apple Macintosh G4 Powerbook computer. 28

38 CRDI Overlay The standard face of the CRDI dial was used in this study with both positive and negative segments listed as shown in Figure 1. Prior to running subjects, the CRDI laminated overlays were affixed to each dial. Figure 1. Standard Continuous Response Digital Interface (CRDI) Dial Procedure Volunteer participants were greeted outside the study room and then asked to take a seat at one of the eight CRDI listening stations. Once seated, participants were asked to read the consent form and instructions, and to fill out the demographic questionnaire (see Appendix B). The questionnaire responses were gathered as subject demographic data for informational purposes only were not considered part of the experimental study. Once the participants 29

39 completed the above, a brief explanation of the CRDI dial was given and participants had a short time to examine and manipulate the dial and ask any necessary questions. The following written instructions were given to each participant: You are going to hear five renditions of Giacomo Puccini s Nessun Dorma from the opera, Turandot. As you listen, you will be asked to move the dial in front of you corresponding to your felt emotional response to the music. Feel free to move the dial as much or as little as you like, utilizing as much or as little of the entire range as you feel is appropriate. Please complete the following and we will begin shortly. Once this portion of the study was completed and all questions answered, participants were asked to put on their headphones and prepare to begin the listening portion. No starting point for the CRDI dial was specified prior to the start of the study or for the beginning of each excerpt; however, most participants moved the dial to the neutral position. It was thought that each participant would quickly move the dial to a corresponding position within a few brief seconds of hearing each excerpt. Once everyone appeared relaxed and comfortable, both the musical stimulus (CD) and CRDI software were activated simultaneously. During the course of the study, approximately ten minutes, participants did not interact with each other. Upon completion of the listening portion of the study and once the participants answered their final exit question, the researcher thanked each participant as they left the study room. 30

40 CHAPTER FOUR RESULTS This study was designed to determine if participants from varied performance-media backgrounds have what they consider a felt emotional response while listening to several renditions of Puccini s Nessun Dorma through performance media the same as and different from their own and to examine the emotional response of each of five participant groups. Specifically, how does this emotional response compare to the emotional response shown through their own performance medium? Secondarily, the study examined whether the response patterns over time of the Continuous Response Digital Interface (CRDI) dial show a relationship between listening groups and/or performance media. Continuous Response Digital Interface (CRDI) responses from participants in five performance media groups (marching band, non-musician, piano, voice, and wind ensemble) were compared as participants listened to five two-minute excerpts of Puccini s Nessun Dorma set in five different performance media (voice, DCI/marching band, solo piano, popular dance music, and wind ensemble). In an effort to control for possible order effect, participants heard one of five different counterbalanced orders. Data were gathered as participants manipulated the dial of the CRDI indicating the magnitude of their emotional response (on a range from 0 to 255) to the music stimuli. Participants responses were sampled twice a second as they listened to each of the five excerpts. At the conclusion of the study, data for each participant was transferred to an Excel file and sorted by participant group. Only the data of participants who answered, yes in the questionnaire to the post study question, Did the movement of the CRDI dial correspond to your felt emotional response to the music? were used in the study, which represented over 95% of the total participants. Means and standard deviation were determined and graphically displayed. Graphs include separate composite means and ongoing standard deviation response graphs for each except and one composite response graph detailing overall mean and overall mean standard deviation by group for each excerpt. Visual inspections of the composite graphs demonstrate both large as well as subtle differences among and between groups. The ongoing changing responses of participants within groups provide the best description of responses. 31

41 A musical footprint based on group emotional response is clearly evident in each of the five excerpts, especially when compared and contrasted next to one another as shown in Figure 2. Further visual investigation required that each excerpt be individually extracted and analyzed separately, allowing for greater visual detail and qualitative analysis. In each excerpt, smaller, more subtle, peaks, valleys, and plateaus in emotional response can be observed. Mean and Standard Deviation Response- All Excerpts Emotional Response Groups Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble 50 0 Voice Excerpt DCI/Marching Excerpt Wind Ensemble Excerpt Time (in.5 second units) Dance/Popular Music Excerpt Piano Excerpt Figure 2. Composite Response Graph - All Excerpts Voice Excerpt The original/voice excerpt composite graph (see Figure 3, p. 33) provides detailed information through ongoing mean and standard deviation of the emotional responses of the five participant groups. Across all five participant groups, an increase in emotional response was shown from the beginning to the end of the excerpt, however, smaller decreases were demonstrated in the marching band group (through the introduction section) and wind ensemble 32

42 group (opening few measures and entrance of the solo vocalist). In each case, the slight decreases were followed by an increase in emotional response, albeit a gradual or a more drastic one. In all groups, the highest emotional response levels or peaks were recorded towards the end of the excerpt, after the peak of the final statement and finale. The highest peak of emotional response was recorded in the wind ensemble group (229), followed by both the piano and voice groups (226). These three participant groups combined represent music majors while the other two participant groups combined (marching band and non-musicians), represent nonmusic majors. The marching band group (212) and the non-musician group (180) demonstrated the lowest peak of emotional response also towards the end and finale of the voice excerpt. In addition, the marching band and non-musician groups also demonstrated the lowest start-tofinish difference in response (+44) while the other three music major groups were led by the highest start-to-finish difference in the voice group (+74), followed by the piano group (+67) and the wind ensemble group (+54). Ongoing standard deviations for each participant group are also presented and graphed in Original/Voice Excerpt Groups Emotional Response Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble 50 End of Introduction Section and Beginning of Peak- Initial Crescendo Initial Crescendo Time (in.5 second units) Peak- Final Statement and Finale Figure 3. Composite Response Graph - Original/Voice Excerpt 33

43 Figure 3. The highest levels of standard deviation were shown in the non-musician group, which was also lower in mean emotional response from start-to-finish. In most cases where high readings of standard deviation are observed, a general lack of agreement in responses is indicated. All other participant groups with the exception of both the marching band and wind ensemble groups stayed fairly consistent in standard deviation from start-to-finish. Both the wind ensemble and marching band groups demonstrated a substantial decrease in standard deviation immediately after the peak of the final statement and finale while reporting at the same time, a substantial increase in mean emotional response. This indicates that as their emotional responses towards the end of the excerpt increased, the agreement within the group also increased. Final observations and visual analysis of the musical contour of each participant group shows what the peak response levels indicated; that the marching band and non-musician groups were indeed lower in their emotional responses than the other three groups despite showing a slight and gradual increase over time. Upon entrance of the solo vocalist within the first few seconds, the piano and voice groups show a spike in response leading to a more gradual increase over time. Surprisingly, the non-musician group does demonstrate a very slight but noticeable increase in emotional response upon entrance of the vocalist immediately followed by a decrease. The other two groups, wind ensemble and marching band, demonstrate a decrease in emotional response upon the entrance of the vocalist. It is also interesting to note the plateau seen in the response of each group over the period of time between the peak of the initial crescendo and the peak of the final statement and finale. However, increases in emotional response can be observed in all groups as the final statement nears its peak. Even more of a drastic spike in emotional response can be observed immediately after the peak of the final statement. DCI/Marching Band Excerpt The DCI/marching band excerpt composite graph (see Figure 4, p. 35) provides detailed information through ongoing mean and standard deviation of the emotional responses of the five participant groups. The highest peak of emotional response recorded was shown in the marching band group (214), near the very end of the excerpt. Slightly lower, the non-musician group (212) reached its peak emotional response at the same point in time. Over the course of the excerpt, the contours of both the marching band group and the non-musician group are very consistent and 34

44 similar to one another. Considerably lower in terms of peak emotional response was the voice group (177) and wind ensemble group (176). The piano group demonstrated the lowest peak response of the five groups (138). With each of these three groups, the peak emotional response occurred right after the peak of the initial crescendo followed by a decrease in emotional response immediately afterwards. Once again, the non-music major and music major groups were similar in their responses in the DCI/marching band excerpt as was seen in the voice excerpt. In all five groups, there was a positive difference in emotional response from start-to-finish, however; there were noticeable differences among the groups. Along with the highest peak emotional response recorded, the marching band group also demonstrated the largest start-to-finish difference within the excerpt (+57), followed by the non-musician group (+51) and the wind ensemble group (+41). Despite the noticeable decrease in response towards the end of the introduction and the subsequent and dramatic increase in emotional response through the peak of the initial crescendo and into the dramatic tempo change, the voice group showed the smallest start-to- DCI/Marching Band Excerpt Groups Emotional Response Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble 50 End of Introduction and Beginning Peak- Initial Dramatic Tempo Peak- Final Statement and of Initial Crescendo Crescendo Change Finale Time (in.5 second units) Figure 4. Composite Response Graph - DCI/Marching Band Excerpt 35

45 finish difference in response (+5). The piano group was only slightly above (+10) yet more consistent across time. The ongoing standard deviation for each group is also indicated in Figure 4. From the dramatic tempo change through the peak of the final statement and finale, both the marching band and non-musician groups demonstrated the lowest levels of standard deviation with the more consistent being the marching band group. From the beginning of the initial crescendo to the finale, both groups showed a gradual increase in emotional response while demonstrating a steady agreement within each group shown through standard deviation. The highest overall levels of standard deviation were observed in the piano group. The piano group demonstrated the highest peak of standard deviation (87) along with the wind ensemble group (87). The lowest peak of standard deviation (44) as well as the lowest valley of standard deviation (24) was found within the marching band group. With final observations and visual analysis of the musical contour of each participant group, several observations can be made. From the start of the excerpt and as previously mentioned, similar overall contours can be seen in the response of the marching band and nonmusician groups. Both groups seemingly plateau from the very start through end of the introduction and show a similar response from the beginning of the initial crescendo to the finale and end of the excerpt. Similarities in response in the marching band, non-musician, wind ensemble, and voice groups from the beginning of the initial crescendo to the dramatic tempo change can also be observed. These groups show a very similar and substantial response to the initial crescendo through its peak with the most drastic emotional response to the crescendo coming from the voice group. The piano group appears to show a consistent level of emotional response across the excerpt with the exception of lesser peaks and valleys around the initial crescendo and its peak. A valley or decrease in emotional response often follows peaks or increases, and this is evident in the emotional response of the piano group. This is also evident in the emotional responses that follow the dramatic tempo change. Upon the dramatic increase in tempo, the three music major groups all demonstrate a small gradual decline in emotional response. At this same musical moment, both the marching band and non-musician groups show a gradual increase towards the end of the excerpt. Surprisingly, the piano group demonstrates a steady increase in emotional response from the peak of the final 36

46 statement and finale to the end of the excerpt; however, at the same time, the piano group also demonstrates its highest levels of standard deviation. Wind Ensemble Excerpt The wind ensemble composite graph, (see Figure 5) in comparison to the other excerpts used in the study, is perhaps the most consistent and similar in contour between the groups. Both ongoing mean emotional response and standard deviation are similar across each of the five participant groups. Peak emotional response levels were comparable with the levels occurring towards the end of the excerpt. The non-musician group demonstrated the highest peak of emotional response (226) followed by the voice group (216), wind ensemble group (215), and both the marching band and piano groups (206). While there are differences, these levels are relatively close to one another in comparison to the other excerpts. In addition to similar mean emotional response contours, group standard deviation emotional responses are among the lowest levels of the five excerpts. The voice group had the lowest valley of standard deviation Wind Ensemble Excerpt Groups Emotional Response Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble 50 End of Introduction and Beginning of Initial Crescendo Peak- Initial Crescendo Peak- Final Statement and Finale Time (in.5 second units) Figure 5. Composite Response Graph - Wind Ensemble Excerpt 37

47 reported (24) while the piano group demonstrated the highest peak level of standard deviation (69). While the contour of the emotional responses is similar, there were noticeable differences in emotional response from start-to-finish. The largest difference in emotional response from start-to-finish was demonstrated in the voice group (+84). The voice group was followed by the start-to-finish differences of the non-musician group (+69), the piano group (+63), and the wind ensemble group (+58). The marching band group demonstrated the smallest amount of start-tofinish difference (+40). The most substantial response to a specific musical moment from all five participant groups was demonstrated during the initial crescendo to its peak. This musical moment represented a unified and parallel increase in emotional response followed by a plateau and slight decrease in emotional response on the other side of the crescendo. The marching band group indicated both an increase and decrease in emotional response through the introduction and was below the other groups in emotional response from that musical point to the end of the excerpt. Popular Dance Music Excerpt Upon initial visual investigation of the popular dance music composite graph, (see Figure 6, p. 39) one can quickly conclude that the both non-music major groups (marching band and non-musician) responded considerably higher in emotional response to the excerpt than the other participant groups. Difference in the peak emotional response of the participant groups was also substantial. The highest peak of emotional response in the popular dance excerpt was demonstrated in the marching band group (173) followed closely by the non-musician group (170). Although these numbers represent high peaks of emotional response in the context of this excerpt, these peaks are low in comparison to the other excerpts. The emotional response peak of the non-musician group occurred immediately after the beginning of the electronic drums while the peak emotional response of the marching band group was reached several times over a period of time after the harmonic key change. The peak emotional responses of the other participant groups (music majors) were demonstrated immediately at the beginning of the excerpt: wind ensemble group (163), voice group (154), and the piano group (150). However, emotional responses succinctly and sharply declined once the excerpt began. 38

48 Popular/Dance Music Excerpt Emotional Response Groups Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble End of Introduction and Electronic Counter Entrance of Main Melody Harmonic Key Change Beginning of Electronic Drums Melody Time (in.5 second units) Figure 6. Composite Response Graph - Popular Dance Music Excerpt Ongoing standard deviation in emotional response also differs between groups. The marching band group demonstrated the lowest level of standard deviation (31) while recording the highest emotional response levels during the introduction. Peak standard deviation levels among the other participant groups were considerably higher than the marching band group with the highest level reading in the non-musician group (85), followed by the piano group (81), the wind ensemble group (83), the voice group (76), and the marching band group (60). With the standard deviation contour being consistently high in the participant groups, with the exception of the marching band group, it shows moderate disagreement in emotional response between participants within groups throughout the excerpt. The popular dance music excerpt was the only excerpt of the five to record a negative start-to-finish difference in emotional response of a participant group, which occured in four of the five groups. The marching band group was the only group to demonstrate a positive difference in emotional response from start-to-finish (+18) while each of the other four groups demonstrated a negative difference, with the voice group recording the largest difference (-63); followed by the wind ensemble group (-58), the piano group (-42), and notably less, the non- 39

49 musician group (-5). The repetitive nature of the excerpt in addition to the lack of a true finale may account for the negative difference in emotional response from start-to-finish. As far as the overall footprint or contour of the excerpt, there appear to be several key observations. While the emotional responses from the participant groups appear to be spread and different, there are several parallel and unified movements as musical moments occur. The very beginning of the excerpt shows both differences and similarities between groups. Four of the five groups immediately demonstrate a decrease in emotional response to the excerpt within the first few seconds, with the exception being the marching band group, which stays consistent. While the three music major groups sharply decrease in their emotional response over the course of the introduction, the two non-music major groups demonstrate a slight increase in their response. Immediately following the end of the introduction and beginning of the electronic drums, there were slight increases and peaks in each of the five groups with the most noticeable spike in emotional response demonstrated by the voice group. However, after this larger increase in the voice group and after the smaller spikes in the other groups, a consistent and unified decrease in emotional response was recorded. This decrease in emotional response in all five groups was followed by an increase in emotional response as the entrance of a syncopated electronic melody was introduced, followed yet again by another decrease in emotional response across all five participant groups. Upon the introduction and entrance of the main melody, the main motif from Nessun Dorma, all five groups showed similar and substantial increases in emotional response leading up to the harmonic key change. Upon the key change and modulation, the voice group demonstrated the only noteworthy movement in emotional response as their response decreased while the other groups either increased slightly or stayed consistent. However, towards the end of the excerpt, the voice group did demonstrate an increase in emotional response while the nonmusician group, along with the piano group, continued a slight increase in emotional response. The emotional response from the wind ensemble group was the most erratic through the final half of the excerpt and finished with a decrease in emotional response towards the end, while their instrumental counterparts in the marching band group virtually stayed the same. 40

50 Piano Excerpt The piano excerpt composite graph (see Figure 7) shows an overall positive difference and increase in emotional response from the beginning to the end of the excerpt. Peak emotional responses were much more varied in contrast to other musical excerpts. The highest peak of emotional response was demonstrated in the wind ensemble group (215) shortly before the peak of the initial crescendo. The piano group, shortly after the peak of the initial crescendo, demonstrated their peak emotional response (195). The peak emotional response of the nonmusician group (193) occurred similarly to that of the wind ensemble group, shortly before the peak of the initial crescendo. The peak emotional response of the voice group (189) occurred during a plateau after the peak of the initial crescendo and before the peak of the final statement and finale. Finally, in terms of peak emotional response levels, the marching band group demonstrated their peak emotional response (170) at the end of the excerpt. Start-to-finish differences within each participant group were fairly close with the exception of the wind ensemble group. As a result of the drastic increase at the start of the Piano Excerpt Emotional Response Groups Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble 50 0 End of Introduction and Beginning of Initial Crescendo Peak- Initial Crescendo Time (in.5 second units) Peak- Finale Statement and Finale Figure 7. Composite Response Graph - Piano Excerpt 41

51 excerpt, the start-to-finish difference in emotional response within the wind ensemble group was quite substantial (+96). Despite the opening seconds of increased emotional response, the wind ensemble did show the highest response levels across the excerpt. The voice group (+41) and piano group (+31) followed the wind ensemble group in start-to-finish emotional response difference, once again grouping the music majors together. Both the marching band (+24) and non-musician (+20) groups demonstrated the lowest start-to-finish emotional response differences. Ongoing standard deviation in responses was varied, especially as seen in the piano group. The piano group had the highest level of standard deviation (93), which occurred near the end of the excerpt. This was notably higher than the standard deviation of the other groups at that same musical moment. This is notable as the mean response from the peak of the initial crescendo toward the end of the excerpt actually declined in the piano group while the standard deviation reached its highest point. This illustrates a moderate lack of agreement within the group as the excerpt neared its finish. The same reverse trend can be seen in the emotional responses of the wind ensemble group, which clearly responds higher to this excerpt than the other groups. From the start, as the emotional response of the wind ensemble group increases through the introduction and initial crescendo, a substantial decrease in standard deviation is also shown during that same time, permitting the assumption that the wind ensemble group was more in agreement in their emotional response to that specific musical moment. Finally, the lowest valley of standard deviation was demonstrated in the voice group (31) as their emotional response increased through the introduction. In terms of standard deviation, the other groups were fairly consistent across time. Overall Mean Response and Overall Mean Standard Deviation Comparison As shown in Figure 8 (p. 43) group emotional responses to the musical excerpts were examined to provide an overall mean response and an overall mean standard deviation. The bar graph provides an efficient and clear manner in which the overall emotional response to all five musical excerpts can be compared to each other. As the musical footprint in each excerpt is different (musical moments happening at different points in time) a standard line graph, as used in the analysis of individual excerpts, was not used. 42

52 Mean Response and Mean Standard Deviation 250 Emotional Response Groups Mean- Marching Band SD- Marching Band Mean- Non-Musician SD- Non-Musician Mean- Piano SD- Piano Mean- Voice SD- Voice Mean- Wind Ensemble SD- Wind Ensemble 0 Voice Excerpt DCI/Marching Excerpt Wind Ensemble Excerpt Excerpts Popular/Dance Excerpt Piano Excerpt Figure 8. Composite Overall Mean and Standard Deviation Response Graph Marching Band Listening Group The marching band group demonstrated their highest overall mean emotional response to the two band instrumental excerpts, the wind ensemble excerpt (181) and the DCI/marching band excerpt (172). The DCI/marching band excerpt did record the lowest overall mean standard deviation (35) of the five excerpts while the wind ensemble excerpt demonstrated the fourth highest overall mean standard deviation (48). This analysis surmises that while the marching band group did not respond the highest to their own performance media excerpt in terms of overall mean emotional response, they did demonstrate the highest level of agreement within the group to the DCI/marching band excerpt. The wind ensemble and DCI/marching band excerpts were followed in overall mean emotional response by the voice excerpt (171), the popular dance music excerpt (163), and the piano excerpt (153). Not only was the piano excerpt lowest in overall mean emotional response, but it also recorded the highest overall mean standard deviation (50) within the marching band participant group. Although the popular dance excerpt was fourth of the five excerpts in overall 43

53 mean emotional response, the marching band group did demonstrate the highest response of the participant groups to that excerpt (163). Non-Musician Listening Group Similar to the marching band group, the non-musician group demonstrated their highest overall mean response to the wind ensemble excerpt (190), followed by the DCI/marching band excerpt (179). Both of the these excerpts also demonstrated the lowest levels of overall mean standard deviation compared to the other excerpts; wind ensemble excerpt (40) and the marching band excerpt (42), indicating a general agreement within the non-musician group. What was most surprising was the low level of overall mean emotional response to the popular dance music excerpt demonstrated by the non-musician group, which was last among the five excerpts (148) for these participants. However, the overall mean standard deviation for the popular dance music excerpt was the highest overall for that excerpt (75) so, while the excerpt received the lowest overall emotional response from the non-musician group, the high level of mean standard deviation indicates the lack of consensus within this group. This overall mean standard deviation (75) was the highest overall mean standard deviation recorded by the five groups for any of the five excerpts. Piano Listening Group The piano group demonstrated their highest overall mean emotional response to the voice excerpt (199) while demonstrating the lowest mean standard deviation (53). The wind ensemble excerpt was their second highest in overall mean emotional response (183) with the second lowest mean standard deviation recorded (57). The piano excerpt was the third highest excerpt in overall mean emotional response (173), grouping the three music major performance media excerpts higher in overall mean emotional response than the non-music major performance media excerpts; the marching band excerpt (123) and the popular dance music excerpt (102). While the piano excerpt was third in mean emotional response, the excerpt was the highest in overall mean standard deviation (72) of the five excerpts, demonstrating that the emotional responses towards their own performance media excerpt did not consistently agree within this group. The mean standard deviation of the piano group (72) was followed in highest to lowest deviation by the popular dance music excerpt (69), the DCI/marching band excerpt 44

54 (68), the wind ensemble excerpt (57), and the voice excerpt (53). The piano group is the only group in the study to record an overall mean above 50 in standard deviation for all excerpts. Voice Listening Group The voice group was one of two groups whose highest overall mean emotional response was recorded during their primary performance media excerpt, the voice excerpt (195). Once more, the music major performance media excerpts comprised the top three excerpts in overall mean emotional response for this group as the wind ensemble excerpt (184) and the piano excerpt (172) followed the voice excerpt. The two non-music major excerpts, the marching band excerpt (144) and the popular dance music excerpt (98), were the lowest in overall mean emotional response. The mean emotional response to the popular dance excerpt (98) represents the lowest mean response demonstrated to any excerpt by any participant group in the study. In terms of overall mean standard deviation, the voice group demonstrated the two lowest results in the study in comparison to any other participant group. On the voice excerpt, their primary performance media, the overall mean emotional response (195) was matched by the second lowest mean standard deviation (35) in the study. The voice group was in agreement in their emotional responses to the voice excerpt. The mean standard deviation for the wind ensemble excerpt (31) was the lowest recorded in the study. While the popular dance music excerpt was lowest of the five excerpts in overall mean emotional response by the voice group, the mean standard deviation was the highest of the five excerpts (66). Both the marching band excerpt (48) and piano excerpt (48) were the next highest excerpts in terms of mean standard deviation. Wind Ensemble Listening Group The overall mean emotional response of the wind ensemble group was highest for its own performance media, the wind ensemble excerpt (192). Further, this mean emotional response for the wind ensemble excerpt occurred while recording their lowest mean standard deviation of the five excerpts (42). The piano excerpt (191) and the voice excerpt (190) followed the wind ensemble excerpt in overall mean emotional response, thus grouping the three music major performance media excerpts at the top in terms of mean emotional response. As shown, the three music major performance media excerpts were each rated notably high and near each other in 45

55 mean response comparison. The lowest results in mean emotional response came from the two non-music major performance media excerpts with the marching band excerpt (151) considerably higher than the popular dance music excerpt (100). The lowest mean standard deviation for this group was demonstrated for the wind ensemble excerpt (42). The next highest mean standard deviation results were found in the other two music major performance media excerpts; the voice excerpt (50), followed by the piano excerpt (54). The marching band excerpt (68) and the popular dance music excerpt (73) followed the music major performance media excerpts in mean standard deviation. Highest Overall Mean Emotional Response As indicated in Figure 8 (p. 43), the overall mean emotional response and overall mean standard deviation of each participant group for each performance media excerpt is provided. The five highest results in overall mean emotional response were demonstrated as follows: the piano listening group for the voice excerpt (199), the voice listening group for the voice excerpt (195), the wind ensemble listening group for the wind ensemble excerpt (192), the wind ensemble listening group for the piano excerpt (191), and a tie between the wind ensemble listening group for the voice excerpt (190) and the non-musician listening group for the wind ensemble excerpt (190). Clearly, the three music major performance media groups and their corresponding excerpts are all listed among the highest in mean emotional response with the exception of the non-musician listening group and their mean response to the wind ensemble excerpt. In two of the cases, the voice listening group and wind ensemble listening group, their response to their own corresponding performance media excerpt were listed together among the highest in mean response results. From the group standpoint, three of the five highest results in overall mean emotional response were from the wind ensemble listening group, with the piano and voice listening groups taking the other top two of the five positions. From the excerpt standpoint, the voice excerpt demonstrated three of the five highest results in overall mean emotional response, the wind ensemble excerpt demonstrated two of the highest, and the piano excerpt had one result. 46

56 Lowest Overall Mean Emotional Response Both the marching band excerpt and the popular dance music excerpt received the lowest ratings of overall mean emotional response. As indicated in Figure 8 (p. 43), the five lowest ratings of overall mean emotional response were demonstrated by the voice listening group for the popular dance music excerpt (98), the wind ensemble listening group for the popular dance music excerpt (100), the piano listening group for the popular dance music excerpt (102), the piano listening group for the DCI/marching band excerpt (123), and the voice listening group for the DCI/marching band excerpt (144). In each case listed, one of the non-music major performance media excerpts, either the popular dance music excerpt or the DCI/marching band excerpt, were rated lowest in terms of overall mean emotional response by each of the three music major participant groups. Specifically, the three lowest ratings of mean emotional response were from the voice, wind ensemble, and piano listening groups for the popular dance music excerpt while the DCI/marching band excerpt followed with its lowest responses. In this case, both of the non-wind instrument music major groups, piano and voice, demonstrated the lowest mean response for the DCI/marching band excerpt. 47

57 CHAPTER FIVE DISCUSSION This study was designed to determine if participants from varied performance-media backgrounds have what they consider a felt emotional response while listening to several renditions of Puccini s Nessun Dorma through performance media the same as and different from their own and to examine the emotional response of each of five participant groups. Specifically, how does this emotional response compare to the emotional response shown through their own performance medium? Secondarily, the study examined whether the response patterns over time of the Continuous Response Digital Interface (CRDI) dial show a relationship between listening groups and/or performance media. Continuous Response Digital Interface (CRDI) responses from participants in five performance media groups (marching band, non-musician, piano, voice, and wind ensemble) were compared as they listened to five two-minute excerpts of Puccini s Nessun Dorma set in five different performance media (voice, DCI/marching band, solo piano, popular dance, and wind ensemble). In an effort to control for possible order effect, participants heard one of five different counterbalanced orders. Data were gathered as participants manipulated the dial of the CRDI indicating the magnitude of their emotional response (on a range from 0 to 255) to the musical stimuli presented through different performance media. Participants responses via the CRDI were sampled twice a second as they listened to each of the five excerpts. Means and standard deviation were determined and graphically displayed. Graphs include separate composite means and ongoing standard deviation response graphs for each excerpt and one composite response graph detailing overall mean and overall mean standard deviation by group for each excerpt. A visual investigation of the individual excerpt composite graphs supports existing evidence that when there was any movement of the CRDI dial there was, indeed, some change in the music that could have stimulated the response and that musical changes could be quite extreme or very subtle: the start of a melody, a slight change in a phrase, a change in volume, the slight increase of a singer, or the beginning of a new orchestral texture (Madsen, Brittin et al., 1993). In this study, the response peaks shown in each composite graph of each 48

58 excerpt resulted from the most abrupt and dramatic changes in the music and that even the smallest peaks were recorded and defined by similar musical events presented on a smaller scale (i.e. vocal or instrumental entrance, change in texture, tempo change, etc.). However, the graphic shape of a musical peak is determined not only by what happens before the highest point, but by what happens after (Adams, 1995). Adams points out that the decreased side of peaks seems to be caused by a lessening of something tension (harmonic, rhythmic, or melodic), volume, texture, tessitura or the inability of some people to sustain the intensity of thrill of extreme high points in musical experience. This musical footprint presented based on group emotional response is clearly evident for each of the five excerpts, especially when compared and contrasted aside one another as shown in Figure 2 (see Appendix A). The primary focus of the study was to determine if participants from varied performancemedia backgrounds have what they consider an aesthetic experience through performance media different from their own. As shown in previous research studies examining aesthetic and emotional response to music, both musicians and non-musicians, music and non-music majors, have been the primary focus populations in most studies involving the CRDI. Madsen, Byrnes, Capperella-Sheldon, and Brittin (1993) investigated musicians and non-musicians aesthetic response to music. Adams (1995) examined the emotional responses to visual and aural conditions of music with musicians and non-musicians. Lychner (1998) focused his study on examining the emotional responses within the context of different aesthetic terminology with music and non-music majors. In one of the earliest studies using the CRDI, Gregory (1994a) examined and compared various group responses in relationship to the instrument studied. The five participant groups and the subsequent sample population were chosen for this study based on the use of the corresponding performance media excerpts originally selected to be used. At the onset of the study and during the design stages, the primary focus was placed on the emotional response from each performance media group. This included one non-musician participant group, whose emotional responses could be compared to the responses of the other four musician groups. Adams (1995) used collective response graphs to support previous research findings indicating that musicians respond substantially the same as non-musicians. As indicated in results and through the composite response graphs, the similarities in the responses of musicians and non-musicians are evident in this study as well. The non-musician group responded substantially the same as the musician groups to the wind ensemble excerpt, the piano 49

59 excerpt, and to a lesser degree, the voice excerpt. However, in the case of the two non-music major excerpts, the non-musician group was notably different in response from the other musician groups with the exception of the marching band group. A design to compare the emotional responses of music and non-music majors was not originally intended, however, as the study progressed, it became apparent that the emotional responses of music majors (voice group, wind ensemble group, and piano group) could be compared to the emotional responses of the non-music major groups (DCI/marching band group and the non-musician group). As results indicate, throughout the study, the marching band and non-musician groups tended to respond similarly to the different performance media excerpts and in contrast to the music major groups. This is especially the case with the responses towards the two performance media excerpts directed towards these two groups, the DCI/marching band excerpt and the popular dance excerpt. While the popular dance music excerpt was not responded to as favorably as was the case in the other excerpts, both groups did respond considerably higher than the other music major groups. Consistent with previous research, individual participants in this study were found to have substantially different emotional responses to the excerpts. These emotional responses on the individual level were varied. Differences in individual response could include basic musical preference including the preference or even lack thereof for their own performance media. As the emotional responses on the individual level were found to be quite different, the emotional responses on the group level were equally diverse in some cases. Analysis shown in the composite response graphs (see Appendix A), indicated that group response was spread and notably different in selected excerpts. The popular dance music excerpt was the most dramatic illustration in difference of group response followed by the DCI/marching band excerpt. These two excerpts, paired together as non-music major performance media excerpts, were markedly different as shown through the composite response graphs. At the same time, the other music major performance media excerpts including the original voice excerpt, the wind ensemble excerpt, and the piano excerpt, were all much more close and similar in response. The most consistent excerpt with the highest mean emotional responses, as shown in Figure 5 and Figure 8 (see Appendix A), was the wind ensemble excerpt, which was remarkably similar in the emotional responses of all five of the participant groups. Of course, the entire aspect of collective graphic response, musician versus non-musician, or by any other grouping as 50

60 opposed to individual responses, should be viewed with caution (Madsen, Brittin et al., 1993, p. 188). The excerpts and musical stimuli used in this study were selected with the help and consultation of a panel of experts. At the onset of the study and during the design stages, approximately ten different performance media renditions were considered, finally narrowing down to the final five excerpts. As the results show, and as is clearly evident in the composite response graphs (see Appendix A), there was a substantial difference between the five excerpts in terms of emotional response. The most consistent excerpt across all five participant groups was the wind ensemble performance media excerpt as shown in Figure 5 (see Appendix A). The wind ensemble excerpt was followed by the voice excerpt and piano excerpt in consistency of collective group response. The composite group emotional responses to the two non-music major performance media excerpts were wide-ranging with the popular dance excerpt demonstrating the highest degree of difference in responses. As shown in the results, a general lack of preference and likeability was evident towards the popular dance music excerpt. Individual laughs were evident during the listening portion of the study to this particular rendition. Alternately, the wind ensemble excerpt finished as arguably the most favorable and highly responded to excerpt of the five, including the original vocal rendition, helping support previous research findings that select listeners prefer instrumental rather than vocal music (LeBlanc & Cote, 1983). More research should be done with different populations and various excerpts. It would be interesting to see if similar results would be observed with different populations, different musical stimuli, and different renditions of the musical stimuli. Further research could examine the role of age and other demographic specifics in performance media studies. Does exposure and familiarity play a notable role, as one would expect? When considering the low level emotional responses towards the popular dance music excerpt in this study, one could further investigate how much of a factor basic musical preference plays into emotional responses towards varying performance media. These are just a few of the questions and areas of research that could be the basis for further research in this field of study. 51

61 APPENDIX A COMPOSITE RESPONSE GRAPHS 52

62 Figure 2. Composite Response Graph - All Excerpts 53

63 Figure 3. Composite Response Graph - Original/Voice Excerpt 54

64 Figure 4. Composite Response Graph - DCI/Marching Band Excerpt 55

65 Figure 5. Composite Response Graph - Wind Ensemble Excerpt 56

66 Figure 6. Composite Response Graph - Popular Dance Music Excerpt 57

67 Figure 7. Composite Response Graph - Piano Excerpt 58

68 Figure 8. Composite Overall Mean and Standard Deviation Response Graph 59