WW? Egg 5—3:: 0-» . ’ . ' ‘ ' ‘ 2:“? ' '. :. .:)¢.;‘,"‘|‘v‘ gawk . ' ‘ . #3:; 3553713, . . ' «.39, "A- ‘ .. .. ‘ $31 . .10; :11. l_ . .Y...‘.l\_’. 3 ." . “$931"; Winngfigcm 5:35» 5:34? 4 o n O . h‘ '3 ”35% Q, o :W’ L n , ‘ -. (5‘33"; ‘ . g . _ ‘3’"1 ~ {055, 7’."- 3 “L: ‘ 3’ ”“5"“ h t ' 1 ‘ ‘ A ‘ ' v " ' ‘ m, 3 — Th, firfii? ‘5'"? 'a’z’amfl’ x1 .93 2%}? ma... W“ _‘~'i=$" .:"' - 235x . flown." .W'. .6.) .c e-vx. “1:34. {“7" 33: fig? .4qu M"":‘t.’ ' ‘ "“‘ *7 . 3,47} ‘ '9, ‘3‘":1 5.1,! "1' 7373151.; fig; 4,. ’ . ~ ‘. ‘ ._ ,' ‘ ' “(5%” “if“? “ " ”é“ "’9: 33"?th “i“ _ 2' 33.54:: 9,, . j- “I?“ L'l‘h'fi, . , _ : ”4' ~ V $3? ‘ \ ,‘~ ‘I 5’??? J ‘ é’afifn ‘ " I 1' 3% 2? '4. v. « M I. ‘E,’:~: 1 7 ii. '91 N .- ‘ ll... a ‘.»\"&d x» J“.._.L 3&1" . ~ 91... .. "N's" “a. H ‘ fir" - . .~ vz‘a“=r5“ as. «m. \‘L:{t€j~k\::§ Rixfflfifl '1 ,2 ‘.. '5 3'4» ,«figfiifim‘fiu 1’- ‘ ~33; .. u _ ,A , , I . .n , d 4 .. ' {355.511- - ' “4‘2":‘wttflflfl; w'ififié‘ "* ‘ m . “W 5' :-----,;r..-.\.egé:.~24~,.» . ~ ' . Q ra‘lfi‘. ‘ . 37‘ A a.’ 1 J Eh ‘ Lc‘ 141.13g. 7‘5 2 3-33.- ~*»'§53‘é'\};’;§{~2-?:2$ “~33: ‘fikficsx Fran " {'- .3». '31.“: 7.1 1-. ' ‘ .xtk,.-.39m. ll'3‘v'ljyfif" 3i: 1. m . a. . t wk 3‘. u. .4. v' 9‘ d: ' i ‘1 n 3:" iv ’ " '19-“ ‘3 _ \ v’LJ.;;~ .. . 3. . ..u\.}m‘§: ' ‘ 1.5“}; -. A . 11> 31;: , -. - M as w- -‘5 a a. . . .. x. . fifufl"; *K-‘I‘y‘. 7" -. ,, . 0.. v V. .7 . .\ ’1‘ 5 \. ‘ 7 . .*.-~.‘15"-“ V~ “5 'i ; .f‘ " t. . ‘-r.:» 53‘. kW ‘ . r 3.. . ‘.. . .7 m: - ~ 1 ’ ‘ " 15;»: , A: ‘ o“ ' "‘ "‘ “A 59312;. . a \ 393%; 3b- ; ‘ 3'3 L}"’:"Zéagi.fit, ' “Y I > “fir-'7» . W4 -~.-,- Few ‘ ' . r: K’ 2,- a; w”! ' A "~ A _ ‘. _ - u’ «’L; JS‘QKJTJ‘Ik." Iv ‘xru ." . “. ”5:34.115 ‘ ‘ ‘ ': «.1 I .-.- P“ . ~. ' ‘ fl . ,‘,-. e‘uv waf' 2".” . A ~ w.- ..' J. 3*”;qu ‘ . c‘. «:55. . :‘I.‘\LV,K . l" as; . 5 J. J 0' . - .g\;é-,‘14 I . ‘ '- ,' 1w , :‘a. m .. fir: r I “a“ 1 r I. 1 II: a’:% a" a 1' ' '75 . m":- a: I ll :- . ', f‘l‘,; q)". 44’, I A ..'.; 'M‘l‘" ’1" ~:"1‘.. WW.- 9 § 1’ l'l I 'rV "‘ 'JV ['7'- v‘ ..v\ 'v 'lr -‘_ ._, “gfiflj' _ "5} +5.. J;.-‘ <- "‘:Y‘ 1+!" v-1 4". .‘T 3 ~‘f‘1 @73345’77 fr“* _‘ Egan-.. .? mum sm. Illlllllll’HlllllllhlllllllllllllllllllHIJIIIHIIIIHIIHI 1 University 31293 00771 2452 \~ I This is to certify that the dissertation entitled The Effects of Listeners' and Performers' Race on Music Preferences and the Relationship Between the Listeners' Expressed Music Preferences and Expressed Preferences for Biack and White Social Encounters presented by Jan H. McCrary has been accepted towards fulfillment of the requirements for Ph.D. degnxin Music Education fl/Zwtfia‘, Major professor Datej/-é " ?0 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE ll RETURN BOX to remove this chockom from your record. TOAVOID FlNESrdunonorbdoroddo we. DATE DUE DATE DUE DATE DUE MSU Is An Aflinnntivo Action/Equal Opportunity Institution THE EFFECTS OF LISTENERS' AND PERFORMERS' RACE ON MUSIC PREFERENCES AND THE RELATIONSHIP BETWEEN THE LISTENERS' EXPRESSED MUSIC PREFERENCES AND EXPRESSED PREFERENCES FOR BLACK AND WHITE SOCIAL ENCOUNTERS BY Jan McCrary A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY School of Music 1990 ABSTRACT THE EFFECTS OF LISTENERS' AND PERFORMERS' RACE ON MUSIC PREFERENCES AND THE RELATIONSHIP BETWEEN THE LISTENERS'EXPRESSED MUSIC PREFERENCES AND EXPRESSED PREFERENCES FOR BLACK AND WHITE SOCIAL ENCOUNTERS BY Jan McCrary Working with 102 middle-school listeners and 119 college listeners, I examined black and white listeners' preferences for music examples by black and white vocal performers and the relationship between the listeners' expressed music preferences and their expressed preferences for social encounters with people of different races. Responding to a prepared listening tape of 20 music examples by black and white performers, the listeners indicated how much they liked or disliked each music example by checking a point along a seven-step Likert-type rating scale anchored by the responses "I Like" and "I Dislike." The listeners were then directed to identify the performer's race by selecting a point closest to the racial identifier that the listener believed accurately identified the performer's race. The racial perception rating scale, also a seven- point Likert-type scale, was anchored by the descriptors "Black" and "White." The respondents' attitudes toward people of different races were examined on a second measurement that included 18 statements describing different social-encounters with blacks and whites. The responses to the social-encounter measurement showed same-race preferences for both the white respondents and the black respondents. But the pattern of responses to the listening measurement was somewhat different. The importance of the performer's race was clearly demonstrated among the black listeners. These listeners' preference for music examples by the black performers and their identification of the black performers' race was strong and positive. Moreover, the black listeners were more successful than the white listeners when identifying the performers' race. When the black listeners identified the performer's race as white, their negative preference ratings demonstrated the black listeners' overall dislike for the white performers. White listeners showed more listener-flexibility in their preference responses. The white listeners' preferences for the black performers were almost equally as high as their preferences for the white performers. Although the middle-school listeners' overall preference for the music examples was low, listener- flexibility for the black and white performers was greater among the middle-school subjects than among the college subjects. Copyright by JAN HALLIBURTON MCCRARY 1990 To my parents. ACKNOWLEDGMENTS I wish to thank Dr. Albert LeBlanc for his patience, guidance, and constant encouragement through the completion of this project. My sincere appreciation is also extended to committee members Dr. Charles McDermid, Dr. Corliss Arnold, Dr. Dale Bonge, and Dr. Lawrence Redd for their assistance toward the completion of the research. vi TABLE OF CONTENTS Chapter Page I. BACKGROUND . . . . . . . . . . . . . . . . . 1 Introduction . . . . . . . . . . . . . . . 1 Purpose . . . . . . . . . . . . . . . . . 8 Problem . . . . . . . . . . . . . . . . . 9 Limitations . . . . . . . . . . . . . . . 10 Assumptions . . . . . . . . . . . . . 10 Need of the Study . . . . . . . . . . . . 10 II. RELATED LITERATURE . . . . . . . . . . . . . 13 A Foundation of Music Preference Research 13 Overview of Related Studies . . 15 Effects of Specific Listener Characteristics 24 The Function of Social Groups: Ethnic and Peer Influences . . . . . . . 28 Approval from Authority Figures . . . . . 41 Other Variables Influencing Affective Response Patterns . . . . . . . . . . 42 III. PROCEDURE . . . . . . . . . . . . . . . . . 47 Introduction . . . . . . . . 47 Development of the Instruments . . . . 48 Acquiring Subjects for the Pilot Study . . 59 Acquiring Subjects for the Final Study . . 63 Data Collection . . . . . . . . 65 IV. RESULTS . . . . . . . . . . . . . . . . . . 72 Data Analysis . . . . . . . 72 Analysis of the Racial- -Encounter Measurement . . . . . . . . 85 The Combined Effects of Race and School- Level . . . . . . . . . . 89 Observing Respondents' Behavior . . . . . 96 Examining the Open-Ended Responses . . . . 98 V. DISCUSSION 0 O O O O O O O O O O O O O O O 102 Conclusions and Recommendations . . . . . 115 APPENDICES Page A. Racial Encounter Measurement . . . . . 119 B. Response Booklet for Listening Measurement . . . . . . . . . . . 125 C. Consent Forms . . . . . . . . . . . . . 135 REFERENCES . . . . . . . . . . . . . . . . . . . 137 Table 10. LIST OF TABLES Order of Music Examples on the Listening Tape Demographic Characteristics of the Sample Black Listeners' Preference and Perception Scores White Listeners' Preference and Perception Scores Preference Scores by Listeners' and Performers' Race O O O O O O O O O O O O O O O O O O O O Mean Scores for Responses to Social Encounters with Blacks and Whites College Listeners' Overall Preference and Perception Scores Middle-School Listeners' Overall Preference and Perception Scores . . . . . . . . . . . . . . Race Perception Scores for Frank Sinatra and Lou Rawls Race Perception Scores for the Different Music Styles Page 57 65 77 78 81 86 91 92 105 109 CHAPTER I Background Introduction In an examination of research on approaches to teaching music listening skills and modifying music preferences, Hedden (1981) states that "one of the goals many music educators attempt is to increase their students' liking of certain kinds of music" (p. 21). In response to this need, music preference research has intensified during the past years. Research questions have addressed the following: what kinds of music and music elements are preferred by listeners, what effect does familiarity exert on music preferences, what are the effects of different teaching strategies on music preferences, and what effects, if any, do social influences have on listeners' preferences (Finnas, 1989). The pioneer work of Hevner provided an impetus for much of the research which followed her studies. For instance, Hevner's adjective circle (1935; later revised by Parnsworth, 1954) has provided researchers with terminology that is sometimes used to describe aesthetic experiences and affective responses to music stimuli. Other examples of 2 early research that shaped the direction of later studies include Schuessler's (1980; publication of his 1947 dissertation) examination of the effect of cultural and racial background on music listening preference. His findings indicate that cultural and sociological factors influence music preferences. Farnsworth (1950) suggests that musical preference is not determined at birth but is culturally derived. Noting surveys of preferences for programming concert literature, Farnsworth explains: "history teaches that within our Western culture area taste [preference] has changed appreciably from time to time" (p. 23). When specifically identifying aspects of one's culture, Parnsworth includes the listener's social group, occupation, and educational background. LeBlanc's (1982) theoretical model for a theory of music preference is predicated on findings of previous research in music preference. In a visual representation of the proposed theory, LeBlanc's model demonstrates that hierarchical levels of variables specified in the model may affect listener preferences. The model also identifies clear distinctions between and among variables that represent characteristics of the listener, variables within the listener's environment, and variables that represent the physical properties of the music stimulus. Each of these variables may work individually and interactively to influence a listener's music preferences. 3 Central to any investigation of music preference is the listener's attitude. A study of attitude requires an examination of one or more components: a respondent's affect for the attitude object, the respondent's observable behavior toward the attitude object, and the respondent's cognitive responses to the attitude object. Summers (1970) identifies four areas which an attitude addresses: "attitude is a predisposition to respond to an object rather than the actual behavior toward such object,...attitude is persistent over time, . . . attitude produces consistency in behavioral outcroppings, . . . attitude has a directional quality" (p.2). Another view is presented by Price (1986), who adopts the following definition for attitude: "a learned predisposition reflecting the way one feels about a subject while not in the presence of that subject, which is not directly observable." It is this "predisposition," combined with affective factors within the environment, characteristics of the listener, and characteristics of the music stimulus which may influence listeners' musical preferences. Earlier studies in the field of attitudinal research examined factors that influence a respondent's predisposition to respond to an attitude object. These studies examined the effects of age, social and cultural background, and race or ethnicity. Sub-populations of children, adults, and social and racial groups were 4 frequently used as research samples. Data collection for the studies was often by self-report, by observation of overt behavior in response to the attitudinal object, or by physiological responses to the attitudinal object or some representation of the object (Wapnick, 1976; Summers, 1970). Studies which have examined the effect of listeners' age on music preference affirm the importance of maturation on attitude development. Meadows (1970/1971), Prince (1972a), Greer, Dorow, and Randall (1974), May (1985), Peery and Peery (1986), Sims (1987), Flowers (1988), and LeBlanc, Colman, McCrary, Sherrill and Malin (1988) have all found that age influences listener preference responses to music. For young listeners of art music, an apparent turning point appears to be about grade four, with overall preferences for young listeners declining through grade seven, but when testing listener preferences for traditional jazz music, LeBlanc, et al., (1988) found that presumably due to maturation, preferences may "rebound" after seventh grade. In the latter study, over 900 subjects were used; the respondents ranged from grade three through college undergraduates. With age as a primary factor under examination, results from the study reveal a U-shaped overall preference curve. Third-grade listeners were the youngest subjects tested and overall preference for the music examples was highest at this grade level. Listener preference began to decline at grade four and the decline 5 continued through grade seven with seventh-grade listeners revealing the lowest preference point. Listener preference for the traditional jazz music examples began to rise again after grade seven and continued to rise through high school, reaching the highest level with college age listeners. When one examines the influence of racial background on listeners' music preferences, research findings indicate that black and white listeners tend to prefer music styles performed by members of their own racial group (Schuessler, 1980; Meadows 1970/1971; Appleton 1970/1971). And indeed, when Appleton (1970/1971) examined the respondents' record buying habits, he found that over 88% of the artists preferred by black respondents were black, although by contrast, he found that just 68% of the artists listed by white respondents were white. Moreover, the responses to the music listening stimuli indicated that the black listeners preferred music examples by black performers, but white listeners, demonstrating stronger preferences for racially differing artists, preferred perfonmances by white artists and some of the black artists. Both Meadows (1970/1971) and Appleton (1970/1971) used music examples that were currently popular during the time of testing, so the music examples were frequently heard through the popular media. Recent investigations using familiar and unfamiliar recordings have found similar results. Behavioral observations reported by LeBlanc (1979), and 6 LeBlanc and Sherrill (1986) indicate that black listeners prefer black-perfonmed music examples while white listeners prefer white-performed music examples. In addition, the studies report that certain styles usually associated with one race prompted ridicule from some listeners of the racially different group. Killian (1990) examined the effect of model characteristics on listener preferences. Exploring parallel research in sports and social psychology, Killian reports that an observer's imitation of a model is based on the observer's perception of any characteristics the model holds which are similar to those of the observer (Bandura & Walters, 1963). The most salient model characteristics are race and gender, and these can have the greatest influence on an observer's choice to imitate the model. In addition, the perceived success or power of the model can strongly influence an observer's imitation response. These imitation responses are reported in studies examining observer preference selection for sports heroes. Although the overall preferred selection for boys is for same gender models, with girls' selection frequently crossing sexual boundaries, the imitation response for boys was shown to be affected by the boy's perception of the models' power or success status. The subject's perception of the model's power or success can explain some selections by the male subjects for female models and the female subjects' 7 selection for both male and female models (Bandura, 1971; Bussey 8 Bandura, 1984). Killian (1990) examined seventh- and eighth-grade subjects' preference selection for popular recording artists. Subjects were given a list of the 21 vocalists who performed brief solos on the popular recording USA for Africa: We Are the World. The subjects indicated how well they liked each performer on a prepared response sheet with rating scales of 0 (low) to 10 (high) for each performer. Subjects were then given a copy of the words to "We are the World." With the words, the performer's name followed in parenthesis for the lines he or she sang on the recording. On the response sheet, the subjects circled the three solos they would prefer to sing while viewing the video presentation of the performers singing "We Are the World." Killian explains that the subjects were able to make their selection based on hearing the performances, watching a video presentation of the performers, and reading the performer's name. Although the performer's racial identity and gender were obvious in the video presentation, there was no verbal or written mention of race or gender during the investigation. Killian found that the black students' overall preference selection was for black performers while the white listeners selected many of the black performers, and some of the white performers. Among male listeners, there 8 was also a strong same-sex preference for the models. But race, Killian reports, seemed to unify the subjects' preference and solo choice selections more than gender. Gordon and Abeles (reported in Abeles, Hoffer & Klotman, 1984) examined subjects' perceptions of the music listening preferences for members of different races and gender. The researchers found that the subjects' selections were strongly influenced by the subjects' race and the intended listener's race. For further study, the researchers suggest that researchers examine the relationship of listener attitudes toward differing groups, including racial groups, and their preferences for music generally associated with those groups. Purpose My purpose in this study was to investigate the music listening preferences of black and white listeners for vocal music examples by black and white performers. In addition, I investigated the relationship between respondents' attitudes for members of different racial groups and respondents' overall preferences for the music examples sung by black and white performers. 9 m Examination of the research problem suggested several subordinate problems. The primary problem of the study was to investigate the music listening preferences of black and white listeners for vocal music examples by black and white performers. The following sub-problems were central to investigating the primary problem: 1. to assess listeners' ability to perceive the performer's racial identity; 2. to assess the listeners' overall preference for the music examples; 3. to assess the relationship between listeners' perception of the performer's racial identity and their preference for the music examples; 4. to assess the effect of listeners' race on their preferences for the music examples; Secondary to the music investigation outlined above was the examination of the relationship between respondents' attitudes towards members of different racial groups and the respondents' overall preference for the music examples by black and white performers. To investigate this secondary problem, I pursued the following sub-problems: 1. to assess the respondents' attitude towards members of different racial groups in certain social settings; 10 2. to assess the relationship between respondents' racial attitudes and their preference for the music examples performed by members of each race. Limitations In this study, I observed the following limitations: 1. The subjects' music preferences and racial attitude responses were obtained through written self-report measures. 2. Subjects for the study were drawn from two age groups only: middle-school students from Lansing, Michigan and college undergraduates from Michigan State University. ssum tions I assumed that the subjects' responses were a true and accurate indicator of their attitudes toward the music listening examples. I also assumed that the subjects' responses were a true and accurate indicator of their attitudes toward members of different racial groups. Need for the Study In 1967, the Tanglewood Symposium, sponsored by the Music Educators National Conference and other supporting groups and foundations, addressed specific areas of concern among music educators. Mark (1978) identified three broad areas addressed at the symposium: desirable ideologies for 11 an emerging postindustrial society, the role of music and other arts for individuals and communities in such a society, and suggestions for realizing these potentials. The Tanglewood Declaration was a summary document ensuing from committee meetings and discussions during the symposium. One of the objectives stated in the document is that music of all periods, styles, forms, and cultures belongs in the curriculum. The musical repertory should be expanded to involve music of our time in its rich variety, including currently popular teen-age music and avant-garde music, American folk music, and the music of other cultures (Choate, 1968, p.139). During a ten-year period following the Tanglewood Symposium, the MusicsgdugatorstournalJ published by Music Educators National Conference for music educators, devoted special issues to teaching the music of black Americans and the music of different ethnic groups. The January 1970 issue of the Music Educators Jouggal_examines teaching the music of black Americans in the classroom. By the September 1980 issue the interests of music education had clearly evolved from black studies to multi-ethnic studies. But research examining both the effect of ethnically-enhanced music curricula on students' musical growth and the need for these kinds of curricula has been sparse. Shehan (1982) reported that no studies existed 12 examining listener preferences for a variety of ethnic musics, and since her observation, only a few studies have emerged (Shehan, 1985; Killian 1987; Darrow, Haack and Kuribayashi 1987). Although the literature indicates that listeners tend to prefer the music of their own ethnic group (Meadows, 1970/1971); Appleton, 1970/1971; LeBlanc, 1979; LeBlanc 8 Sherrill, 1986), studies investigating other variables affecting listeners' preference decisions for ethnically different music are needed. CHAPTER II Related Literature A Foundation for Music Preference Research When evaluating listener responses to music, early experimental researchers were faced with distinguishing between the philosophy of the aesthetic experience and the psychological domain of affective responses to music stimuli. Hevner (1935), in an attempt to define precisely the nature of musical expression and musical meaning, borrows from aesthetic philosophy when identifying tension and relaxation in musical expression central in the human response to music. Hevner (1936) identifies musical meaning as being "the affective value and the expressiveness of music" (p. 246). More precisely, Hevner specifies musical events related to rhythm, melody, and harmony as elements which may be primary enhancers of musical expression. After examining the writings of Hevner and others on the human response to music, Abeles (1980) offers this distinction between the aesthetic experience and the affective response in music listening: 13 14 The aesthetic experience [is] an intense, subjective, personal experience, which provides insight into the nature of human life. The aesthetic experience is thought to include some mood, emotional or feelingful aspect, that is, an affective component. The affective response is generally conceived of as a more superficial response than the aesthetic experience (p. 105). But the experimental research of Hevner lends even more credence to the study of affective responses to music stimulii: her research findings pinpoint certain patterns in individual responses to specific music examples. Acknowledging that the human response to music requires a certain "flexibility," that is to say, that all individuals do not respond in precisely the same manner to a selected musical composition, Hevner's early findings showed that different listeners' written self-reports to the same composition were indeed similar. These findings led to the development of Hevner's adjective circle--descriptors used by listeners in response to art music examples. According to Bevner (1935), the adjective groups cover "eight well defined affective states or moods" based on listeners' responses to "many varieties of music, long and short examples, in major and minor, and with different rhythms, tempos, and harmonic structures” (p. 191). The results revealed an overall uniformity and consistency among listener responses. As possible influences on listener responses, 15 Hevner (1935) identified characteristics of the listeners which influence their responses to musical expression. Hevner's studies moved beyond merely identifying a listener's musical training as a source of variation in music preferences. 'As listener characteristics, she includes variations in listeners' moods, current affective states, physiological well- being, backgrounds, their personalities, and attitudes toward the music stimuli. In this chapter, I examine research that identifies specific characteristics of the listener, input from the listener's environment, and certain characteristics of music stimuli as sources of influence on music preferences. In addition, a discussion of related attitude research from the social sciences is presented. Qverview of Related Studies In an attempt to define and clarify terminology used in studies exploring affective response patterns to music, members of the 1982 Affective Response Special Research Interest Group (SRIG) of Music Educators National Conference began discussions which eventually led to a proposed glossary for use by music education researchers. The compiled list and definitions are all based on summations of studies and 16 writings by experts in the field. As stated earlier, an aesthetic experience refers to the experience of feelings as a result of perceiving certain musical events. Although research exploring affective response patterns examines listener reactions involving feelings and emotions, the research is also concerned with "learned behavior resulting from a life history of interactions with musical stimuli" (Price, 1986, p. 152). Affective responses encompass "mood-emotional, preference, and taste responses." Farnsworth's (1950, 1969) writings focus on musical taste and preference. Musical taste is the "overall attitudinal set one has toward the phenomena which collectively comprise music" (Farnsworth, 1969, p. 97). Parnsworth explains: "the attitudes built up in oneself toward composers and toward the modal, finality, key, and other musical effect--all quite clearly form a part of one's musical taste" (p. 97). A person's "musical taste" is commonly believed to represent a long-term commitment and yet, from individual to individual, musical taste will vary as much as other behaviors shaped by social pressures. Although Parnsworth's (1950) writings focus on musical taste, his studies seem to examine both long-term commitments to music phenomena and statements of music preference--"short-term, less stable affective 17 responses" (Abeles, 1980, p. 110). Farnsworth believes that musical taste is not innate but culturally derived. He states: We hardly need proof to be certain that taste develops out of experiences in our homes, churches, clubs, and schools, and out of experiences with the concert stage, record disks, the radio, and the printed page. . . . We come to have several standards of taste: for the concert stage, for the dance hall, for church, and for school. . . . Age, intelligence, and special training are important variables in this process of taste formation. (p. 66). This shows agreement with the early writings of Hevner in acknowledging that listeners' preferences for musical selections are based on many variables. Farnsworth also recognizes the listener's environment, the music itself, and characteristics of the listener as sources of variation in musical tastes and preferences. Prince's (1972) paradigm demonstrates interrelationships among variables that influence listeners' affective response patterns. Prince states that a listener is "capable of more than one response pattern and more than one type of listening experience" (p.446). He believes that individuals are capable of listening to the same or different music examples and of responding differently to the examples because of differences in listening situations. Prince's paradigm was proposed to assist future researchers in their 18 attempt to isolate possible variables as sources of variation in listener response patterns. Again, this is consistent with Kevner's early findings. When listeners responded to specific music examples, their selection of descriptors was consistent overall. But Hevner reports that variations existed in the subjects' responses, varying from strong listener agreement, to listener descriptors representing completely opposite responses. LeBlanc's (1982) proposed model for a theory of music preference includes three broad areas to categorize specific variables that influence music preference: the characteristics of the listener, characteristics of the music stimulus, and input from the listener's environment. These variables have been either tested or suggested as sources of influence in research studies on music preference. The model not only identifies specific variables in each of the broad categories but graphically shows possible relationships and interrelationships among the variables and their positions within different levels of the model's hierarchy. These writings and other studies by music researchers cited in this section are consistent with and often based on established attitude research in social psychology. Wapnick (1976) explains that 19 preference research in music evolved from the early attitudinal research in social psychology. Early studies examined subjects' attitudes toward "many different populations in varying situations" (p. 1). Investigators, by and large, examined the attitudes of children, adults, social groups, and racial groups. Data collection was often by self-report, observation of the subject's overt behavior in response to the attitudinal object, or the subject's physiological responses when in the presence of the object or some representation of the object (Summers, 1970). With this research, procedures were devised for sampling large populations, and measurement scales were developed to assist in measuring the intensity of respondents' attitudes (Wapnick, 1976). The study of attitude toward music or other attitudinal objects requires an examination of at least one of the following attitude components: the respondent's affect toward the object, the respondent's observable behavior when in the presence of the attitudinal object, and the respondent's cognitive responses toward the object. These components specifically identify the "evaluative" element in an attitude, the "intentional" element in an attitude, and beliefs the individual holds about the attitude object (Rajecki, 1982). Cook and Selltiz (1964) explain that 20 an attitude cannot be measured directly, but must be inferred from observable behavior "whether the behavior be language in which the individual reports his feelings about the attitude-object, performance of a task involving material related to the object,...or actions toward a representative of the object-class" (p.37). Music researchers have used written and verbal self-reports, behavioral observations as well as respondents' reports of behavioral intent, and physiological measures to evaluate listeners' music attitudes. One of the earliest definitions for "attitude" adopted in social psychology was proposed by Allport (1935). He stated that an attitude is a "mental or neural state of readiness, organized through experience, exerting a directive or dynamic influence upon the individual's response to all objects and situations with which it is related" (cited in Rajecki, 1982, p.4.). Summers (1970) identifies the following parameters of attitude: an attitude is a predisposition to respond to an object rather than the actual behavior toward such object; attitude is persistent over time; attitude produces consistency in behavioral outcroppings; attitude has a directional quality (p. 2). Music education researchers have adopted the 21 following definition for attitude: it is "a learned predisposition reflecting the way one feels about a subject while not in the presence of that subject, which is not directly observable" (Price, 1986, p.152). Each of these definitions identifies the directional quality of attitudes, which can also be measured in terms of intensity. In addition, the consistency with which the individual responds to the attitudinal object is highlighted in each definition. Ajzen and Pishbein (1977) believe that "attitudes are held with some aspect of the individual's world," such as another person or a physical object within the individual's environment (p.889). Although an attitude may not always predict a given behavior, the researchers believe that an individual's overall pattern of responses to an attitudinal object is directly influenced by that person's attitude. Simply stated, "attitudes are held and behaviors are performed" with respect to certain attitudinal and behavioral objects (p. 889). When the strength of the correlation is low between the intention to perform a given behavior and the actual behavior, then attitudes and behaviors are sometimes inconsistent. Fishbein and Ajzen identify four elements of attitude and behavior "entities" or objects that can influence attitude- behavior relations: 22 the action, the target at which the action is directed, the context in which the action is performed, and the time at which it is performed (p. 889 . An individual's responses may vary because of the attitude and behavioral elements and any interrelation among them. A strong relationship between attitudes and behaviors exists when the correspondence between the elements is high. The authors explain that, at least, the relation between the target and action elements of the attitude and behavior must be strong. Attitude measurements are used as predictors of behaviors. Researchers using these measures as predictors assume that there is a strong degree of consistency between the respondent's evaluation of or affect held for an attitude object and his or her actual behavior when in the presence of the object. Music researchers have objectively evaluated student attitudes for music styles or music activities with physiological measures, self-report measures, and behavioral measures (Kuhn, 1980). Kuhn found that self-report measures are quite efficient and adequate when examining music preferences. These can include open-ended questions, paired comparison techniques, multiple choice scales, pictographic scales, rating scales measuring the intensity of the subjects' attitudes, summated rating scales, and semantic 23 differential scales. Music educators are frequently interested in how to increase students' preferences for certain kinds of music (Kuhn, 1980; Hedden, 1981). Attitude measurements are often used in studies that examine aspects of the listener, his or her environment, or specifics of the music stimulus and the influences on listeners' preferences. Hedden's (1981) review of research results, and research reviews by Wapnick (1976) and Abeles (1980) present studies that use attitude measurements as predictors of certain musical behaviors. Comprehensive discussions of music preference have emerged since Wapnick's (1976) timely review of literature on music attitudes (Abeles, 1980; Hedden 1981; Finnas, 1989). But it was Prince (1972) who advanced a paradigm formalizing the study of music preference that systematically identifies possible influences on music attitudes. LeBlanc (1982) proposed a theory of music preference which visually demonstrates possible relations and interrelations of variables that may affect listeners' preferences within a model's hierarchy. LeBlanc later enhanced the model by indentifying broad categories of variables that are consistent with previous research studies (LeBlanc & Sherrill, 1986). The broad categories for the 24 variables are the characteristics of the listener, the listener's cultural environment, and input information from the music stimulus itself. Effect_ of4§pecific Listener Characteristics Some investigations of specific listener characteristics have examined the effect of the respondent's ethnic or racial identification (Schuessler, 1947; Meadows, 1970/1971; Appleton, 1970/1971, Madsen 8 Madsen, 1975; Darrow, Haack & Kuribayashi, 1987; Killian 1990). Other investigations of listener characteristics have tested the effect of maturation on listeners' music attitudes (Bauman, 1960; Meadows, 1970/1971; Prince, 1972b; Greer, Dorow s Randall, 1974; May, 1985; Sims, 1987; LeBlanc, Colman, McCrary, Sherrill & Malin 1988). Greer, Dorow 8 Randall (1974), compared pre-school through sixth-grade students' preferences for rock music and non-rock music examples. Students' listening time for non-rock examples increased from the pre- school through first- grade listeners and reached the highest amount of listening time with the third-grade listeners. Listening time for non-rock examples decreased at grade four and continued to decrease through grade six. But the average listening time reported for the rock music examples increased steadily 25 rom nursery school listeners through grade 6. The older students showed the greatest preferences for the rock music examples. The authors report that the time between third and fourth grade "appears to be a pivotal time in terms of musical taste" (p. 289). May (1985) later reported similar findings: older students in his study, who were in grade three, had overall preference responses leaning toward the popular music styles. May's (1985) and Sims' (1987) findings confirmed that with younger subjects, specific music preference inclinations are established by grade four. May examined first- through third-grade students' preferences for musical styles and found the younger children more tolerant of different musical styles. Similar to findings by Greer et al. (1974), May reports that with older listeners, the overall preference was for the popular music examples. Sims (1987) examined pre-school through fourth- grade listeners' tempo preferences for classical piano music. She reports that preference decisions based on specific music characteristics--tempo--are probably established by grade four. This study confirms findings by LeBlanc and McCrary (1983) who state that listeners tend to prefer faster tempos. But, Sims' study also supports research that examined the effect of age on children's music preferences; she found that 26 subjects' overall favorable attitude responses for the classical music examples decreased with advancing age. The pre-school subjects demonstrated the highest preference responses while the fourth-grade subjects gave the lowest preference responses to the music stimulus. Similar findings are reported by Flowers (1988), who also reports a significant decrease in positive responses between fourth- and fifth-grade respondents and fifth- and sixth-grade respondents listening to art music examples. LeBlanc, Colman, McCrary, Sherrill, and Malin (1988) examined different-aged listeners' preferences for different levels of tempo. The study, using traditional jazz music examples, expanded an earlier study (LeBlanc & McCrary, 1983) by examining the effect of maturation on listeners' preferences for different tempos. The subjects in the study ranged from third- grade through college undergraduates. Again, the pivotal point for the listeners appeared to be between grade three and five. Overall preferences, for all tempos, were greater for third-grade listeners, the youngest grade-level tested. But overall preference responses began to decrease with the fifth-grade listeners. The results represent a U-shaped curve: the strongest preferences were reported by the youngest listeners and the oldest listeners. After grade three, 27 for each advancing grade level through grade seven, preferences continued to decline. The lowest overall preference rating was reported for seventh-grade listeners. After grade seven, preferences began to increase with each advancing grade level, re- establishing a high preference point with the college- age listeners. These studies indicate that age is a significant listener characteristic influencing preferences for popular and non-popular music examples. Before fourth- grade, young listeners seem to respond positively to certain non-popular music examples. But grade four emerges as a pivotal point with the listeners, and current popular music becomes the preferred style. Positive responses for non-popular examples decreases from fourth grade through early adolescence and a rebound effect seems to occur during later adolescence. Positive responses continue to increase for non-popular music examples into the early college years. Radocy and Boyle (1979) state that preference decisions are frequently based on societal pressures: "a person making a musical choice considers opinions of other persons who are significant in his or her life, as well as cultural messages in and about the music" (p. 231). LeBlanc's (1982) model specifically identifies these factors as input from the listener's 28 cultural environment and includes peers, educators, family members, and authority figures. Although one's ethnic or racial identity is a listener characteristic, other members of the racial or ethnic group provide environmental contributions as peer and cultural influences. Often, the music's cultural messages address ethnic, age, racial, religious, and other sub- populations within the cultural environmental. The Function pf Social Groups: Ethnic and Peer Influences Pestinger, Schachter, and Back (1950) examined the effects of peer group influence in social settings. Central to the investigation was the formation and influence of small groups on individual conformity in a housing community. The researchers found that small face-to-face groups form on the basis of common sex, common occupation, or common experiences. Further, these smaller groups tend to exert greater influences on individual conformity than larger groups. The effectiveness of the group depends to a major extent on how "cohesive” the group is (p.11). Within the group, attaining prestige, social status and the approval of others are the goals of the individual. This results in a high degree of uniformity within the group. Pestinger, et al. (p. 170) explain that there is a 29 "resulting similarity in the behavior of members of the same group," and from this evolves the development of a group standard. The group standard includes "the acceptance of a given pattern of behavior based upon a given set of attitudes and idealogies for all members of the group" (p. 170). The group uses this standard to encourage conformity from new members and the group standard becomes self-reinforcing for others within the group. Similar factors influence the behavior of individuals in a cultural setting and can be extended to ethnic communities that exihibit strong cultural ties. Festinger et al. (1950) believe the customs and institutionalized patterns of the cultural environment act as strong factors that influence social behavioral patterns of its members. The individual's attraction to the group is founded on the common experiences of group members and the type and degree of contact between members. This attraction to the group, combined with the goals of the group, can influence the responses of the individual. When influenced by a listener's race, music preference decisions are based on cultural input from the racial group or its ethnic sub-group and not on the biological differences that distinguish different racial groups. Aboud and Skerry (1984) define an 30 ethnic group as a "socially and/or psychologically defined set of people who share a common culture or cultural background, often because of similarity of race, nationality, or religion" (p. 3). Further, the researchers define an ethnic attitude as "a predisposition to respond in a favorable or unfavorable manner toward people from different ethnic groups" (p. 3). Aboud and Skerry (1984) provide a critical review of literature that addresses the development of ethnic attitudes. Because these authors identify racial similarity as one descriptor for an ethnic group, their review of literature is appropriate to the discussion of music preference decisions based on race or ethnicity. Aboud and Skerry examine literature addressing the development of own-group and other-group attitudes in children. The literature suggests that own-group preferences of whites emerge by age four or five. But after age seven or eight, studies show a decrease in own-group preferences among whites. Some studies report a decline well into adolescence while others report no change in own-group preferences after age seven or eight. Black children expressed a preference for whites as early as age four, but in some studies, the subjects 31 in this age group were equally divided between those preferring whites and those preferring blacks. As white own-group preferences decreased between age five and seven, black own-group preferences increased. Aboud and Skerry conclude that there tends to be an inverse relationship between the intensity of attitudes held toward one's own group and attitudes held for other groups, and the perceptions among children and adults of their own-group and an other- group. As an individual becomes more preoccupied with him or herself as a group member, the differences among other-group members are evaluated by the individual. Through time, the individual accentuates between-group differences and within-group similarities (p. 29). The writers explain that "the more similar one rated one's own group, the more dissimilar one rated the other groups, and both ratings increased with age" (p. 21). Two findings are consistently reported in Aboud and Skerry's review: four-year-old subjects appear to demonstrate more consensus in their ethnic attitudes, supporting the belief that attitudes emerge around age four while black subjects developed an increase in own- group preferences and a decrease in other-group preferences with each advancing year after age four or five. Doyle, Beaudet and Aboud (1988) report that as 32 subjects' ages increase, there is an increase in ”ethnic flexibility." In the study, the researchers explain that older children provided both positive and negative adjectives to their own and to a different ethnic group. Although the subjects in this study were Canadian, the results support findings reported in the literature review of Aboud and Skerry (1984). Researchers in music education have examined the effect of listeners' racial backgrounds as a primary variable, and as a secondary variable with other main effects. Primary in Schuessler's (1980; publication of his 1947 dissertation) investigation was the effect of subjects' socio-economic-status on music preference decisions. In the study, Schuessler also examines black and white female adolescents' music preference responses. The results of his study indicate that preference decisions were based on the subjects' racial differences, but the author did not report the differences in the responses as decisions of preference. Although the study examined the cultural input of racially different adolescents, Schuessler did not recognize that the cultural differences may have contributed to the differences in the subjects' expressed preferences. Meadows (1970/1971) examined the influence of race, among other variables, on music preferences. The 33 music stimuli included several musical styles to elicit listeners' preference responses. Meadows reports that black respondents preferred jazz, blues, "soul," and spirituals, while white listeners preferred rock, country, "classical," folk and Broadway showtunes. The listeners' preference responses for the performers clearly mirrored the listener's racial identity. Appleton (1970/1971) examined the preference responses of black and white college students to different styles of music. In the study, Appleton defines the music used for the stimuli as "folk" and "popular" styles. Similarly to Meadows, Appleton's findings demonstrate strong listener preferences for music performed by members of their own racial group. Measuring behavioral intent, Appleton requested that subjects list the recording artists whose records they most often purchase. Over 89% of the artists listed by the black respondents were black while 68% of the artists listed by white respondents were white. Although the responses to the music listening stimuli show strong preferences of the black respondents for black performers, white respondents' preference responses were extremely positive toward both the "soul," a form of black popular music, and rock music examples. This finding confirms ethnic attitude research reported by Aboud and Skerry (1984), and 34 Doyle, Beaudet and Aboud (1988): older respondents demonstrate more ethnic flexibility; whites' other- group preferences tend to increase with age; and blacks other-group preferences decrease with age. Killian (1990) reports similar findings, with white subjects demonstrating more racial flexibility in their preference responses than black subjects. Killian (1990) examined seventh- and eighth-grade subjects' preference selection for popular recording artists. The subjects' selections were based on hearing the performances, watching the performers during a video presentation of the music example, and reading the performer's name. The artists' racial identity was obvious to the listeners, although no verbal or written mention of race was made by the researcher during testing. Killian reports that the black students' overall preference selection was for black performers while the white listeners selected many of the black performers and some of the white performers. Other variables examined in the study included age and gender, but race unified the subjects' responses more than any other variable. Still other researchers have found race and ethnicity prominent in determining listeners' music preferences. Madsen and Madsen (1975) examined the preference responses of sixth-grade students from a 35 predominantly black school. The musical choices were "traditional music" frmm elementary music series recordings and "soul" music. The students consistently selected the "soul" music selections, a style of popular music by black Americans, over the traditional music selections. In LeBlanc's studies, (LeBlanc, 1979; LeBlanc 8 Sherrill, 1986) information from behavioral observations by the examiners show a preference by black listeners for black performed music examples and a preference by white listeners for white performed music examples. In addition, both the 1979 and 1986 study report that certain styles prompted ridicule from some listeners from the other racial group. In his original paper, LeBlanc (1979a) reported observations of the listeners' behavioral responses to specific listening examples. White listeners demonstrated unfavorable overt behavior toward a music listening example which was performed in traditional black gospel style. This finding was deleted in the editorial process in the publication of the same study (LeBlanc, 1979b). Similar findings were reported in LeBlanc 8 Sherrill (1986). Again, members of racially different groups demonstrated unfavorable responses to certain music examples: black listeners toward certain white performances, and white listeners toward certain 36 black performances. Results from these studies confirm Festinger, Schachter, and Back's (1950) discussion of the influence of social and cultural groups on group members and member conformity to standards established and reinforced by the group. Moreover, the preference studies demonstrate that attitudes toward music-- listeners' affective, behavioral and cognitive responses toward music-~may be strongly influenced by the cultural environment of the racial or ethnic comunity. Darrow, Kaack, and Kuribayashi (1987) state that ethnocentrism--"the tendency to judge other cultures by the standards of one's own"--contributes to cultural biases found in listeners' affective response patterns to the music of different racial and ethnic groups (p. 238). Investigating Japanese and American students' preferences for Eastern and Western music examples, Darrow, et al. report that both the American and Japanese students preferred the Western music examples. The authors explain that the Japanese students' exposure to Western music has been frequent and consistent while the American students have had very little, if any, exposure to Eastern music styles. Further, Japanese students' exposure to Eastern and formal Japnese music is limited and generally through 37 special training, so their liking responses reflect the influence of Western cultures. Group influences on listener preferences can extend beyond racial and ethnic groups to other social and cultural groups (Radocy 8 Boyle, 1979; LeBlanc, 1982). Prince (1972a) identifies this variable in his paradigm as "socially-educationally derived attitudes toward music" (p. 447). Inglefield (1972), Radocy (1975), Killian and Kostka (in press), Furman and Duke (1987), have examined the effect of peer approval on listener affective responses to music. Teacher approval and the effect of other authority figures have been the subject of music preference studies by Greer, Dorow, Wachhaus and White (1973) Radocy (1976), Sims (1986), and Flowers (1988). Alpert (1982) examined peer, teacher and disk jockey approval on listener preferences. Using a measurement to examine personality types, Inglefield (1972) found that respondents conformed more in the presence of their most important "social leaders" than the "rebel leaders," identified as the least important members within the peer-influence groups (p. 64). Inglefield contends that several variables operate in the conformity behavior of listeners and believes that musical preference may be determined by the following personality variables: the 38 listener's need for social approval, his or her relative independence, and inner-other direction. Inglefield's findings demonstrate that the expressed music preferences of adolescent music listeners may be the "result of conformity behavior induced by peer group pressure" (p. 57). Festinger, Schachter and Back (p. 170) believe that "groups have power to impose conforming behavior on members in accordance with the attractions of the members to the group." In Inglefield's study, the "rebel leaders" are identified as the least important members of the group. These would function as group "deviates" in the Festinger et al. theory of group structure. The behavior shown by these individuals is seen as strange and deviant so they are not recognized as part of the group. Other members who wish to "belong" to the group will conform in attitude, behavior, and ideas with the leaders who embody the group's attractiveness. Festinger et al. explain that three group-oriented goals function on member conformity: prestige, social status, and the approval of others. These goals, whether inherent in the group or incidentally mediated by the group, since they add to the attractiveness of the group, also add to its power. Such strong motivations toward belonging to groups enable the group to have a great deal of influence over its members (p.3). 39 Radocy (1976) reports that college music majors conformed to the incorrect stated responses of their peers when examined for individual conformity to group behavior. When members provided erroneous responses to questions relative to pitch and loudness discriminations, the music majors went against their own judgments and conformed to their peers' incorrectly stated responses. But Webster and Hamilton (1981/1982) found that fourth through sixth-grade students were not significantly influenced by peers' music preference ratings; however, in this study, other musical effects were examined. The findings did indicate a tendency for the means to be higher for the positively- influenced peer group and slightly lower for the negatively-influenced peer group. In two separate investigations, Furman and Duke (1988) examined student conformity behavior to verbally stated group preferences in response to hearing original and altered versions of popular music and two versions of orchestra music. There was little change in the original preference statements of the group examined for popular music preferences. The subjects liked the original version which was also the version frequently heard through the popular media. This seems to indicate that familiarity may be operating when subjects are responding to frequently heard popular 40 music. The stated preferences of peers had little effect on the subjects' decisions. However, the preferences of group members exposed to original and altered versions of orchestral music were affected by the contrived preference statements by confederates-- subjects recruited by the researcher to provide predetermined responses to the test items. The test subjects were unaware of the confederate subjects' alliance with the researcher. Killian and Kostka (in press) examined third- through eighth-grade students' conformity behavior to stated preferences for holiday music. The subjects were asked to list their favorite holiday songs and were later shown a list of the "most popular" songs selected by the entire class. Afterwards, the students were again asked to list their favorite holiday songs, but were also given the name of a tune that the researcher said had received the most class votes. In reality, the tune had received very few votes. Killian and Kostka report that many of the songs listed during the first survey were re-named during the second survey, but subjects added new songs to the second list based on the "information" the subjects were given about their peers' selection. Moreover, on the second survey, older students--grades five through eight-- selected the erroneously stated "favorite song" 41 significantly more than did the younger students. From these studies, one can infer that the music preferences of younger children may be influenced less by their peers' judgments. In contrast, the studies indicate a tendency for college-aged students and adolescents to conform to the judgments of their peers. Approval from Authority Figures When examining the effects of authority figures and teacher approval on affective responses toward music, findings are similar but not always conclusive. Sims (1986) reports that subjects' preferences were not influenced by teacher affect in her study. She suggests that perhaps the subjects' age--they were preschool children-~and the short time frame for executing the investigation may have affected the results of the study. But subjects in her study responded positively during music listening activities, particularly when the teacher exhibited high nonverbal affect. Radocy (1976) reports that subjects conformed with the biased judgments of the authority figure. But in his study, the subjects' preference decisions were not always in agreement with their expressed judgments. Findings by Greer, Dorow, Wachhaus and White (1973) indicate that respondents' music selection behaviors were influenced by high approval authority figures. 42 chep Vapiables Influencing Affective Response Patterns In addition to specific characteristics of the listeners and the listeners' environment, music researchers have examined listeners' familiarity with specific music examples and the influence of this familiarity on preferences. Bartlett (1973) examined the effect of repeated listenings on listeners' affective responses and found that attitudes toward the repeated music were improved. Manipulating the tempo and pitch of familiar popular pieces, Geringer and Madsen (1987) found that subjects preferred the unaltered (and therefore more familiar) music examples. Shehan (1985) reports that familiarity through repeated listening combined with participatory classroom experiences positively influenced subjects' affective responses to non-Western music genres. Although the subjects' affective responses toward the taught music selections was positive, there was no transfer of listeners' preferences to the untaught music selections within the same ethnic genres. These studies demonstrate the influence of familiarity on listeners' music preferences. But Wapnick (1976) cautions that conclusions drawn from studies on repeated hearings and the change in attitudes toward music may imply an additional effect. He asserts that "there may be an optimal saturation 43 point beyond which repeated hearings may reverse the effects of earlier positive attitude changes toward the music" (Po 10)- Perhaps the most powerful variable influencing listeners' music preferences is style. When testing listener preference for generic styles, LeBlanc (1979) found that listeners' preferences within the popular music category were greater for easy-listening popular music than for rock. Later studies in the LeBlanc series found that the tempo of the piece, the performing medium, the performer's and the listener's gender, and the listener's age influenced respondents' preferences for music styles (LeBlanc 8 Cote, 1983; LeBlanc 8 McCrary 1983; LeBlanc 8 Sherrill, 1986; LeBlanc, Colman, McCrary, Sherrill, 8 Malin, 1988). Geringer (1982) investigated the effects of age and musical training on listeners' verbal and operant preferences for different music styles. Populations sampled for this study included college music-majors and non-music majors, and elementary students. Using a music listening device which allows the subject freedom of choice over pre-recorded music selections, the listeners selected music examples representative of both traditional composers and popular music composers. Geringer reports that the music majors selected the traditional composers more than the popular composers 44 during their listening time, while the non-music majors and elementary subjects selected the popular composers significantly more than the traditional composers. When researchers report that young listeners tend to prefer popular music styles, the results are not surprising (Greer, Douglas, Dorow 8 Randall, 1974; Shehan, 1985). But when listener preferences for unfamiliar styles of music are examined, the studies reported in this literature review indicate that a listener's preference for the music is influenced by the listener's exposure to the music, the teaching method, certain characteristics of the music such as tempo, characteristics of the listener such as age and gender, and the influences of racial, ethnic and peer groups. Following the Tanglewood Symposium in 1967, many music educators began promoting the study and use of musics of other cultures. One purpose of the symposium was to address the need for new directions in music education. Of central importance was concern for the changing social environment and the effect of this on music education curricula (Choate, 1968). For many music educators, the goal, at minimum, is to broaden students' preferences to include art music. For other educators, the goal is to increase students' preferences for art music and the music of other 45 ethnic, racial and cultural groups (Kuhn, 1980; Hedden, 1981; Shehan, 1985). Music appreciation and general music courses are frequently targeted for middle school students and college age students. LeBlanc, et al. (1988) report that following a decline in overall preferences, positive affective responses may increase after grade seven and continue to increase through college age listeners. In this study, I compare the music preferences of middle school students and college undergraduates. Studies reported in this literature review examining the effect of the listener's racial background suggest that some listeners' preferences may be ethnocentrically based and that different patterns of responses may exist between racial groups. As suggested by the research in social psychology (Aboud 8 Skerry, 1984; Doyle et a1, 1988), the effects of maturation and race may affect the music attitudes of listeners toward racially different music performances. In this study, I examine the effect of the listener's race and the performer's race on listeners' expressed music preferences. Ajzen and Fishbein (1977) believe that "attitudes are held with some aspect of the individual's world" (p. 889). The listener's preferences for racially 46 different performance styles may be influenced by the overall attitudes the listener holds for members of the other race. In this study, using a written self-report measurement, I examine black and white listeners' expressed preferences for music examples by black and white vocal performers. In addition, using a written self-report measurement, I examine the listeners' attitudes toward members of their own racial group and members of another racial group and possible relationships with their expressed music preferences. Central to this study is the evaluation of the effects of race on listeners' music preferences. Modifying the listening-flexibility and music preferences of racially different students' for less familiar styles, requires an examination of the effects of the listener's race and the performer's race on the individual's willingness to attend to music performances. The previous discussion of attitude research in the social sciences and music education demonstrates the importance of examining the effects of race and ethnicity on music attitude development. CHAPTER I I I Procedure Introduction Music researchers have drawn from the attitude research in social psychology when measuring music attitudes. To meet their needs, music researchers have developed and modified instrumentation needed for measuring music attitudes. Notwithstanding, these researchers have been consistent in modeling their own instruments after the measurements developed by attitude research experts in the social sciences. Kuhn (1980), describing the instruments available to music researchers interested in attitude research, outlines physiological, behavioral, and many self-report measures. In his discussion of self-report measures, Kuhn explains that rating scales have been used frequently to examine the intensity of subjects' attitudes. Typically, the subject is required to indicate the degree of his or her own preference or attitude by selecting a point along a continuum reflecting that attitude (p. 9). LeBlanc (1984) states that music preference research usually requires that the subject respond to a music stimulus using one of the instruments available to "measure" 47 48 the strength of the subject's attitude. LeBlanc explains: To measure music preference, researchers must define it for the purposes of their own study, select some aspect of observable human behavior in order to measure preference, collect data which measures the observable behavior, and usually transform the data into quantitative values for statistical analysis. In this study, my purpose was to examine the music preferences of black and white subjects for music by black and white performers. I examined the subjects' ability to accurately perceive the racial identity of the performer to determine if a relationship existed between the subjects' music preference selections and the subjects' perceptions of a performer's race. I also examined the subjects' social attitudes toward members of their own and a different racial group to determine possible relationships between the subjects' expressed music preferences and racial attitudes. This section describes the development of the instruments used for measuring the subjects' responses to the music stimulus and assessing the subjects' responses to attitude statements regarding various racial encounters. The process of acquiring subjects and collecting data are described through the time line from the pilot study through the final study. Develgpipg the Social Encounter Measurement In preparation for conducting the study, I consulted with experts on attitude development and explored scholarly 49 writings on attitude scale development in music education and social psychology. Kuhn (1980) outlines the instrumentation available for measuring music attitudes and LeBlanc (1984) examines criteria that should be used when selecting a response mode for measuring music preferences. In social psychology, writings on attitude scale development by Edwards and Kilpatrick (1948a, 1948b) and Shaw and Wright (1967) were useful, and the discussions of Cook, Stuart and Selltiz (1964) and Selltiz, Jahoda, Deutsch and Cook (1966) provided information specific to the development of self- report measurements. In these discussions I found the material I needed to proceed with the development of the attitude measurement. After consulting with a small sample of the population I intended to examine--college undergraduates and middle- school students--I determined that shopping malls represented a social setting familiar to both age groups. Creating attitude statements that centered on face-to-face encounters in a shopping mall setting with racially different people would be appropriate for examining the racial attitudes of both middle-school and college undergraduates. Since direct observation of the subjects' behavioral responses in a mall setting was not feasible, I determined that the measurement should examine the intended behavior of the subjects in the social environment of this 50 setting with racially differing friends, acquaintances, and mall employees. I observed adolescents in shopping malls to determine the actual behavior of the targeted population in mall settings. On other occasions, I initiated and observed discussions of race relations with members of the targeted age and racial groups. From these observations, I created a large number of attitude statements relevant to current expressions of racial attitudes in the described social setting. With outside readers, I examined each statement to determine if each statement clearly represented favorable or unfavorable attitudes. With a panel of writing experts, I examined each statement for its clarity, its appropriateness for the reading levels of the age groups intended, and each statement's sensitivity to the attitude dimension. Items which did not meet the stated criteria were discarded until 18 items remained that represented favorable and unfavorable attitudes toward members of black and white racial groups, and that reflected the social behaviors shared by college students and middle-school students. I administered the measurement to a small group of college students and young adolescents and found that some felt the questionnaire's focus on racial attitudes was too obvious. An outside expert advised that the measurement's focus could be made less obvious by including statements that examined subjects' attitudes toward members of different age 51 groups within the same social setting. Six additional statements, examining the subjects' responses to members of different age groups, were included to dilute the racial focus of the measurement. With the additional statements, the measurement totaled 24 attitude items: 18 measuring attitudes toward members of different racial groups and 6 measuring attitudes toward members of different age groups. The 18 statements of different racial encounters were randomly placed on the attitude measurement and the 6 additional statements were interspersed with the other 18 items. The statements examined to what degree the subject would or would not interact in various encounters with black and white friends, acquaintances and mall personnel. The additional age-related statements suggested similar social encounters, but with members of different age groups. The selected response mode was a seven-point Likert-type rating scale anchored by the responses "I would not" and "I would." The subject was required to select a point along the continuum that reflected the strength of his or her attitude indicating to what degree the subject would participate in the social encounter (see Appendix A). Developing the Listening Measuremgnt Once the attitude measurement was completed, I began preparation for the listening measurement. Before sampling music for the music stimulus, I established criteria for 52 music examples to be selected for the listening tape. I decided to use vocal music recordings, controlling for music style, performer sex, and race. Within each style category, two black performers, one male and the other female, and two white performers, one male and the other female, would be drawn from a pool of music examples. Styles of music with readily available samples of both black and white vocalists included jazz, popular, folk and art music. I listened to representative recordings from each style within different performance periods. White performers were judged for demonstrating vocal stylings dependent more on Western European traditions and showing little or no dependence on indigenous forms of black American music. Black performers were judged for demonstrating vocal stylings dependent more on indigenous forms of black American music described in scholarly writings by Southern (1971, 1983) and Nettl (1976). Gaining access to vocal recordings of art music by both male and female black American performers who demonstrate vocal stylings dependent more on black American singing traditions and less on Western European art music traditions was difficult. Within the art music tradition, the performer typically minimizes vocal stylings that characterize any ethnic or racial group influence. This standard of performance for art music can be found in performances by black Americans and members of other ethnic 53 groups. After listening to several performances by black Americans in the art music tradition, I decided to eliminate art music as a possible category. I then directed my attention toward styles that readily presented vocal stylings by black performers in the stated tradition: gospel, jazz, popular and folk music. While listening to music examples by white performers in these music styles, I determined if the artist minimized the influence of black American vocal stylings and maximized stylings influenced by the Western European vocal music tradition. Admittedly, in some styles--for example jazz-- the vocalist typically demonstrates an influence of black American vocal stylings, thus an attempt was made to select recordings by white performers who demonstrated comparatively less of this influence. The final selection of music examples used in this study was based on the performer's use of vocal stylings traditionally associated with black-American and Western European performance traditions. These samples demonstrated the performer's sensitivity to stylistic interpretations and performance practices related to cultural traditions, and the use of cultural dialect when appropriate. More often than not, the racially different performers demonstrated what I detemmined were subtle distinctions in vocal timbre. To help determine the validity of the listening measurement, I included two music examples, one by a white 54 male and the other by a white female vocalist, whose performances were highly dependent on the influence of black American vocal traditions. In addition, I included two music examples, one by a black male and the other by a black female vocalist, whose performances represented a strong influence of Western European vocal traditions. I included these examples to assess the listeners' ability to accurately perceive the racial identity of the performer as either black or white, even when that performer was singing in the style that showed a strong influence of the other race. To control for style, I selected two recordings by black performers and two recordings by white performers in each style category. To control for gender, of the two black performers selected in each style, one was male and the other female. Of the white performers selected in each style, one was male and the other female. For the validity checks, I selected two examples by black performers who might be perceived by the listener as white: one male and the other female. Also included were two examples by white performers who might be perceived by the listener as black: one male and the other female. I played segments of the selected music examples for musicians and non-musicians who indicated their perceptions of the vocalists' ability to meet the performance criteria within each style category and racial group. I listened to the selected music examples to 55 make a final determination of the suitability of the performer's vocal styling, and to determine the appropriate length of each music example for the music listening tape. The response booklet was designed to elicit responses using a Likert-type rating scale for each music example. For each music example, three responses were required (see Appendix B). First, the subject indicated how much he or she liked or disliked the music example by checking a point along a seven-step continuum anchored by the responses "I dislike" and "I like." The second response required that the subject indicate his or her perception of the vocalist's race. The second scale, a seven-point Likert-type rating scale, was anchored by the responses "black" and "white." The subject was directed to select the point closest to the racial identifier that he or she believed accurately identified the performer's race. A third response was added for each music example to lessen the racial focus of the listening test. Consistent with the attitude measurement, the subject indicated the approximate age of the performer by selecting a point along a seven-point continuum that he or she believed correctly identified the performer's age. The mid-point of the continuum was identified in the response booklet with "age 30" and the scale was anchored with "younger" for younger than 30 and "older" for older than thirty. To elicit open-ended responses, the back page 56 of the booklet requested that the subjects indicate what they liked and disliked about the music examples. Demographic information--the respondent's age, school name, grade level, sex, and race--was requested separately on the front cover of the answer booklet. This information was followed by a written example and a response example for the practice music example on the listening tape. To assist the students, I prepared written directions leading the subjects through the practice example included on the listening tape. I decided to add the directions to the listening tape so that the directions would be heard the same way for each class. Recognizing that my own race could be an unintended factor affecting listeners' responses, I alternated with another speaker-~a white male with a distinguished regional speaking manner. The directions instructed the listener to indicate how much he or she liked the music example by selecting a point along the continuum that reflected the preference. The subject was instructed to select the middle point if he or she was not sure. The subject was then instructed to try to identify the performer's race by selecting a point along the continuum that best reflected his or her perception of the performer's race. Again, if the subject was not sure, the directions instructed the subject to select the middle point. Finally, the subject was instructed to try to identify the performer's age by selecting a point along the Table 1 9_r_der of Music was on the Listggg’ Tapg Dura- Performer Title Race Sex Label umber ticn 1. Frank "Very Good Year" Reprise FS-1025 46 Sinatra 2. Rosetta "What are They Black BL-9625 35 Tharpe Doing in Label Heaven Now" 3. Rita "Am I Blue" A 8 M SP-4531 32 Coolidge 4. Jack ."St. Janes Time/Life S'l'L J03 52 Teagarden* Infirnary 5. Sally "Johnny Sands" Flying FF-409 28 Rogers Fish 6. Big Bill "All I Got Everest FS-213 38 Brocnzy Belongs to You" 7. Judy "Send in the Elektra/ 8E-6002 32 Collins Clowns Asylun 8. Randy "Street Life" KIA MCA-274 40 Crawford 8 the Crusaders 9. Louis "Caldcnia" KIA KIA-274 36 Jordan 10. dnrley "Just Betwea'l RCA LSP-4223 37 Pride* You and Me" 11. Lemtyne "Right as the RCA ARL1-1029 52 Price* Rain" 12. Lou Rawls "Very Good Year" Capitol 84-2948 33 Table 1 (cmt'd.). 58 Dura- Performer Title Race Sex Label Nurber ticn 13. Anita "Farther Alcng" W WCRD SL-6895 33 Bryant 14. Ton Lehrer "Fight Fiercely W Reprise RS-6216 34 Harvard" 15. Rickie Lee "Easy Haley" W Warner Bat-3296 40 Jcnes"t 16. Sarah "Misty" B Mainstreun MRL-2401 42 Vaughan 17. Odetta "If I Had a B Vanguard VSD-43/44 28 Hanmer" 18. Jim Ringer "Going Away" W Folk-Legacy PSI-47 38 19. The Stanley "Rank Strangers" W Gusto SLP-953 41 Brothers 20. Willian M. "Wings" B Elektra/ Lin-114 33 O'Neil with Everest Harlan Christian Tabernacle Choir NOTE: Duration times indicated by total nurber of secaids. * Exalple included as a validity check. 59 continuum that best represented the performer's age at the time of the performance. The taped directions led the subjects through the music practice example, and directed the subjects through the written practice example in the response booklet. The final listening tape began with the spoken directions accompanying a practice music example. Following the directions were 20 randomly placed music examples from four different style categories: jazz, popular, folk and gospel. Table 1 shows the name of the performing artist, the artist's race and sex, the title of the music selection and the selection's placement on the listening tape. An average of 10 seconds of silence concluded each music example, allowing the listener enough time to respond to each of the three response items: the preference response, the performer's perceived race, and the performer's perceived age. The music phrase determined the length of each example. Agguiring Subjects for theAPilot Study I submitted a proposal for review by the University Committee on Research Involving Human Subjects at Michigan State University during March, 1989. Approximately three weeks after the proposal had been submitted, I received notification that the application would be passed on to a second reviewer and was informed that the racial content of 60 the proposal had raised some concerns with the primary reviewer. After some time, the review committee stated that a few changes in the testing procedure and the content of the parent and student consent forms would be necessary before testing began. I made the specific changes requested: clearly stating in the consent forms that responses would be held anonymous, clearly informing the parents and subjects of the racial content in the questionnaire (see Appendix C). The application was re- submitted to the committee with a written statment explaining that I would not test students enrolled in classes I taught. To correlate each subject's responses to the attitude measurement and the listening measurement, I planned to use identification cards. A previously written number on the card would be transferred by the student to the listening measurement and the attitude measurement; the two would be administered at different times. To ensure that the student used the same identification number for both measurements, the student would write his or her name on the card that would be redistributed during the second testing period. The committee was concerned that the proposal did not explain exactly who would have access to the identification cards once the student's name was written on the cards. In the revision, I outlined the procedure that I would use, explaining that I would not have access to the cards after 61 the students' names were written on them. The revised proposal was submitted and approval was granted May 24, 1989. Because of the delay, I decided to conduct a pilot study during the Summer session of 1989 instead of the Spring session. I approached the Department and Area Chair in the Department of American Thought and Language at Michigan State University at the end of the Spring term. Because the American Thought and Language (ATL) deparment serves all undergraduate students at Michigan State University, I could obtain a representative sample of black and white students who were enrolled in ATL courses. The racial content of the attitude measurement was a concern of one ATL area chair. Therefore, he did not grant approval for the use of students enrolled in his area's sections. The chair's program included many minority students and this decision limited the pool of black subjects available for the pilot study. However, approval was granted by other area chairs and I proceeded by requesting permission to test the students of the remaining instructors teaching Summer courses in the department. Two professors responded immediately to my requests and I began testing their students immediately. The total number of subjects used in the pilot study was 35: 77% reported their race as white, 23% reported their race as either black, Hispanic or Asian. The listening measurement was administered to the 62 subjects first, and after a two-week period the attitude measurement was administered. Although observations were made on all data, the primary purpose of the pilot study was to examine the listener's ability to accurately perceive the performer's race. A descriptive analysis indicated that, overall, the listeners were able to perceive the performers' racial identity. But the listeners were not in complete agreement when perceiving the racial identities of black performers Bill Broonzy, Louis Jordan, and Lou Rawls. The listeners were most successful when perceiving the white performers' race. Because of the large sample of white respondents, the results were not surprising. Listeners were instructed to provide written constructed responses after completing the listening test. Analysis of the open-ended responses indicated that the listener's ability to perceive the performer's race was often dependent on the musical style of the performance. One listener wrote "country music is too easy to guess the race" while another wrote "I think I could tell which songs were sung by black people" and continues by citing musical styles she or he recognized as black performance styles. During August 1989, the pilot study's results were submitted to be considered for presentation at the research poster session of the National Convention of the Music Educators National Conference to be held in Washington, D.C., in March 1990. 63 Acquiring Subjects for the Final Study The total number of subjects required for completion of the final project was 200. Because the content of the study depended on the racial identity of the listener, I determined that the racial division of the total sample should be approximately half. Of the total number of subjects, a minimum of 75 should represent black subjects and a minimum of 75 should represent white subjects. In the study, I planned to examine the effect of age differences on the subjects' responses. Approximately half of the total population would be selected from a middle-school and the other half from college undergraduates. With the cooperation of faculty from the Department of American Thought and Language, testing was completed during the Fall term at Michigan State University for the remaining college sample. Seven intact classes were sampled yielding a total of 119 college undergraduates. I submitted an application for permission to test students to the Lansing School District's Office of Research and Evaluation Services August 29, 1989. Approval was granted October 20, 1989. With the notification, I was informed that only one school had agreed to participate. I had selected the Lansing School District because of its close proximity to my home and because the racial mixture of most of the schools was suitable for the needs of the study. I contacted the school's principal after receiving 64 notification from the Office of Research and Evaluation Services. The principal responded to my letter and phone call within two weeks. The principal requested that I give him an opportunity to speak with the teachers and get their responses. A few weeks passed after my initial inquiry and I became concerned about the time lag. I began to inquire with other school districts about their policy for obtaining subjects and made initial steps to select another school district. But in December 1989, the Lansing school principal provided me with the names of teachers who agreed to have their classes participate in the study. In January 1990, I began testing the middle-school subjects and completed examining this age group in March 1990. Six intact classes were sampled with a total of 102 middle school students ranging from grades six through eight. Table 2 shows the demographic make-up of the college and middle-school subjects. 65 Table 2 Demogpaphic Chargcteristics of the Sample. Age Group p_ Total Race Black White Other Middle-school 24 54 24 102 College 56 48 15 119 Gender Male Female Middle-school 46 56 102 College 47 72 119 Note: Listeners who coded their race as Asian, Hispanic, Native American Indian are included under "Other." Data Collection For the final study, 221 subjects were acquired. Of that total, 80 subjects reported their racial identity as black, 102 reported their racial identity as white, 24 reported their racial identity as Hispanic and 15 reported their racial identity as "other,' stating that they were members of Asian or Native American groups. Between the age groups, 102 were middle school subjects and 119 were college undergraduates. Table 2 describes the race and sex 66 characterstics of the middle-school and college subjects. After obtaining administrative approval, I contacted individual classroom teachers who had agreed to participate in the study. I explained to each teacher that I would test intact classes and described what the students' participation required. I showed the instructors copies of the measurements and answered any questions they had. To minimize the effect of my own racial identity on the students' responses to the attitude measurement, I asked each teacher if he or she would be willing to administer the attitude measurement, explaining that the decision was not binding nor the participation required. The teacher decided if he or she felt comfortable administering the attitude measurement or if the teacher preferred that I administer the measurement. Only one middle-school instructor felt uncomfortable administering the attitude measurement. Since the teacher's racial identity was the same as my own, we agreed that my participation should have little influence on the subjects' responses. None of the instructors showed any concern regarding the study's racial content. For the middle-school subjects, I made an initial classroom visit to describe the study to the students and request that each student deliver the parent consent forms home, obtain a parent or guardian signature, and return the form to the classroom teacher the next day. Additional consent forms were left with each classroom teacher for 67 students who needed them. The classroom teachers were instructed to seal the signed consent forms in an envelope and return the sealed envelope to me for filing. Consent fomms for the college subjects were distributed on the first day of testing. Each student who agreed to participate in the study returned the signed consent form before the actual testing began. The signed consent forms were collected and sealed in an envelope by a student volunteer, and then returned to me for filing. For each middle-school and college class, the listening test was administered first, and the attitude measurement was administered approximately two weeks later. Once again, I explained the nature of the test and explained what was required of the subjects. Students were allowed to give their verbal dissent if they chose not to participate in the study. Identification cards, with a number inscribed on the card, were passed to the subjects by a student volunteer. The response booklet for the listening test was then distributed to the subjects and subjects were instructed to transfer the number on the card to the space provided on the answer booklet. I explained to each class that this would ensure anonymity of responses to both measurements. The subjects were then instructed to write their names on the identification cards only or some other identifier that the classroom instructor would recognize. The identification 68 cards were collected by the student volunteer and placed in an envelope and sealed for safe-keeping by the instructor until the second measurement was adminitered. At my direction, the students provided written responses to the demographic section of the response booklet. Middle-school subjects sometimes needed assistance when responding to the question of ethnicity. Many were confused with the generic identifiers of "white" and "black" stating that they were German, or Dutch, or African or some other group that identified them more specifically. Once the demographic data were completed, the subjects were instructed to listen carefully to the tape for further directions. The listening tape was played on a large LaSonic portable stereo cassette player, model TRC-920. The listening tape directed the subjects through the recorded practice example and the written examples in the response booklet. Before the actual listening examples began, I turned off the tape player and asked the subjects if they had any questions or concerns. For both age levels, the directions were apparently clear. The average time for administering the test was 35 minutes. Overall, the college subjects seemed to enjoy the challenge of trying to identify the performer's age and race. Verbal responses were discouraged, but some subjects in both age groups responded with finger snapping, laughter and moans indicating their affective responses to some of 69 the music selections. For example, the Louis Jordan selection elicited laughter from both age and racial groups. The lyrics for this example are humorous and the subjects' responses were directed toward the performer's intended humor. Classes with predominantly black subjects were most vocal. The black listeners showed more positive responses toward the styles that they perceived were by black performers. In predominantly white classes, listeners showed a tolerance for the gospel music examples by black performers, but not necessarily liking-responses for these examples. Racially mixed classes were conservative with their overt behavior: white listeners sometimes giggled shyly during the black-performer gospel examples, and black listeners giggled at the white-performer example by Sally Rogers. Across races and age levels, positive overt responses were observed for the black-performed examples by Louis Jordan and Randy Crawford. In general, the younger listeners' overt behavior showed more dislike for the slow music examples than did the college-age listeners. The classroom teachers were instructed to administer the attitude measurement two weeks after the listening measurement was given. The sealed envelope of identification cards, copies of the attitude measurement, and directions for administering the measurement and transferring the identification numbers to the measurement 70 were left with the instructor. Most of the college professors decided to keep the information in their personal mail-boxes. Other teachers found safe places for the materials. The teachers were told to have each student destroy the identification card once the number had been transferred to the attitude measurement. For absent students, the teacher was given the option to administer the test at a different time, or simply disregard their numbers and only test the subjects who were present. Most disregarded the absent students. Others opted to give the test to small groups of students later. I administered the attitude measurement to two middle- school classes for the teacher who preferred not to administer the test. During the testing, I observed that the students commented about the racial intent of the questionnaire. In the first class, one student asked if the test was "racist." At that point, I briefly stated to the class that each person should give his or her most immediate response to the statements, explaining that the response would represent how each person felt about the statement. This seemed to relieve the concerned student. The total time for administering the attitude measurement was approximately 15 minutes. I later spoke with some of the teachers who had administered the attitude measurement, asking them if they had any problems when administering the test or if the 71 students commented about the content of the attitude statements. The college teachers said that they had not encountered any problems when administering the test. Some teachers stated that the students wondered about the purpose of the test, but none appeared concerned about the measurement's content. But some problems were encountered when one teacher administered the attitude measurement to middle-school students. The racial make-up of the middle-school not only included many black and white students, but other ethnic groups that were well represented at the school. Two of the classes tested had a large hispanic population and some of the students felt that the measurement ignored their own concerns and feelings. Verbal discussions between the students escalated during testing, and the teacher had to dispense with testing one of the classes. After holding classroom discussions that included the teacher and one of the school's counselors, the teacher waited a couple of weeks longer before re-administering the test. I observed that one subject's responses from the class were incomplete and the subject simply refused to answer most of the questionnaire. The incomplete responses were coded zero for the data analysis. Because the subject's ethnicity was reported as "Hispanic," the responses had little influence on the intended analysis. CHAPTER IV Results Data Analysis The raw data for each subject's responses to the listening measurement and the attitude measurement were typed on a Zenith, model 159-12 personal computer. Four copies of the data were saved on 5 1/4 inch disks as ASCII files. I completed the data analyses with SPSS/PC+. Using the statistical program, I assigned variable labels for the response items, and where appropriate, assigned value labels identifying the option-selections for the responses. Running MODIFY VARS REORDER, I reorganized the listening responses into four contiguous areas representing style categories: jazz, popular, folk, and gospel style responses, and a separate category for responses to the validity checks. I performed descriptive analyses for responses to each example in the style categories and conducted reliability analyses using the RELIABILITY program of SPSS/PC+. To determine the reliability of the sub-tests and the total test, I used Cronbach's alpha, an internal consistency approach for measuring reliability. To measure the reliability of responses to the performer's perceived race, I re-grouped the perceived-race 72 73 response items into two categories: black performers and white performers. The same reliability procedure, computation of Cronbach's coefficient alpha, was selected to determine the consistency of the subjects' responses when attempting to identify the performers' race. All preference responses were regrouped, similar to the perception responses of the listening measurement: preference response items for the black performers were grouped together, as were the preference responses for the white performers. Descriptive analyses were performed for the following sub-groups: black listeners, white listeners, college listeners, and middle-school listeners. I observed mean scores and standard deviations of the sub-groups' preferences for the white performers, the black performers, and for the validity checks. I then observed mean scores and standard deviations of the sub-groups' perceptions of the white performers' racial identities, the black performers' racial identities, and the listeners' perceptions of the performers' race included as validity checks. I conducted analyses to determine the statistical significance of the relationship between the preference responses and the perceived-race responses for the following sub-groups: black respondents, white respondents, college- age respondents, middle-school respondents, black college respondents, white college respondents, black middle-school 74 respondents, and white middle-school respondents. To assess the effects of race and school-level, separate tests were conducted for each sub-group on response items for the black performers and for the white performers. Additional analyses were performed to examine the statistical significance of the differences between listeners' ability to accurately perceive the racial identity of the white performers and the black performers. I also examined differences between the listeners' preference responses for the black performers and the preference responses for the white performers. To assess the effects of race and school- level, the analyses were performed separately for each sub- group. Analysis of the four validity examples was done separately, and these music examples were omitted when doing other analyses of the preference and perception scores. Correlation analyses were carried out with the non- parametric Spearman technique. The analyses were conducted to examine the statistical significance of relationships between specific characteristics of the listener and the response items. The non-parametric Wilcoxon signed-rank test was used to examine differences. I computed coefficient alpha to assess the internal consistency of the music listening measurement and the written attitude measurement which focused on racial encounters. For the overall listening measurement, coefficient alpha was .89. Because the total test consisted 75 of four general style categories that equally represented black, white, male and female performers, I computed coefficient alpha for each of the four style subtests. The four style categories produced the following alphas: jazz = .76, gospel = .54, folk = .73, and popular ‘ .52. The following alphas were produced for the perception of the performers' race: white performers = .68, black performers = .51 and the overall test = .57. A problem identified in this study was to assess the listeners' ability to accurately identify the performers' race. Overall; the listeners identified the white performers' race more successfully than the black performers' race. The listeners' perception scores ranged from 1.00 to 7.00: a rating of 1.00 identified the performers' race as black, and a rating of 7.00 identified the performers' race as white. The overall mean score for all of the listeners' perception of the black performers' race was 2.46, and 6.01 for the white performers. Tables 3 and 4 present the mean scores and standard deviations for the preference and perception responses of both the black and white listeners. The black respondents' scores are reported in Table 3, and in Table 4, the white respondents scores are reported. When perceiving the performers' race, the black listeners were more successful than the white listeners. The mean scores and standard deviations indicate that all 76 respondents achieved greater accuracy when perceiving the race of the white performers, although the black respondents attained even greater accuracy when perceiving the white performers' race. The listeners were less successful when perceiving the black performers' race, but, again, the black listeners were more successful than the white listeners. 77 Table 3 Black Listeners' Preference and Perception Scores (n=80) Preference Perception Performer M_ §Q_ M_ §2_ Black Performers ' Rosetta Tharpe 5.16 1.72 1.18 0.79 Big Bill Broonzy 2.51 1.80 2.85 2.33 Randy Crawford 5.58 1.83 1.16 0.79 Louis Jordan 4.41 2.10 2.04 1.87 Lou Rawls 4.16 2.11 2.34 2.25 Sarah Vaughan 3.71 2.19 1.80 1.66 Odetta 2.69 1.92 4.75 2.58 William O'Neil 5.35 1.91 1.15 0.66 White Performers Frank Sinatra 2.76 1.88 6.60 1.15 Rita Coolidge 3.67 1.97 3.61 2.29 Sally Rogers 1.58 1.10 6.86 0.52 Judy Collins 3.26 2.07 6.67 1.03 Anita Bryant 1.88 1.38 6.17 1.65 Tom Lehrer 1.60 1.18 6.75 0.79 Jim Ringer 1.79 1.40 6.66 0.81 Stanley Brothers 1.77 1.54 6.91 0.40 Note: Preference range: 1.00 = low; 7.00 = high. Perception range: 1.00 = black performers; 7.00 = white performers. 78 Table 4 White Listenegp' Preference and Perception Scores (n=102) Preference Perception Performer M _SA_ M SD Black Performers Rosetta Tharpe 2.91 1.74 1.50 1.32 Big Bill Broonzy 2.53 1.60 2.53 1.92 Randy Crawford 3.55 2.00 1.61 1.18 Louis Jordan 4.23 1.96 2.70 1.95 Lou Rawls 3.50 1.91 3.66 2.46 Sarah Vaughan 2.47 1.85 2.54 1.82 Odetta 2.63 1.93 4.30 2.41 William O'Neil 2.27 1.71 1.64 1.42 White Performers Frank Sinatra 3.14 1.93 5.92 1.67 Rita Coolidge 2.94 1.76 4.48 2.26 Sally Rogers 2.70 2.06 6.71 1.04 Judy Collins 4.33 2.17 6.17 1.57 Anita Bryant 2.18 1.32 4.73 2.23 Tom Lehrer 2.74 1.99 6.26 1.36 Jim Ringer 3.01 2.01 5.96 1.66 Stanley Brothers 2.58 1.92 6.35 1.32 Note: Perception range: performer. 1.00 Preference ratings' range: 1.00 low ; 7.00 = high. black performer; 7.00 = white 79 For all of the respondents, perception scores for the validity examples yielded a mean score of 4.59 and a standard deviation of 1.17. When identifying Charley Pride's race, a black performer singing in traditional country style, the following mean scores were yielded: a mean score of 6.33 (§Q_= 1.44) for the white listeners, and a mean score of 6.83 (§Q_= .95) for the black subjects. The scores show that most of the listeners were certain that Charley Pride's racial identity was white and among the listeners, there was strong agreement--especially among the black listeners. Between the two racial groups of listeners, the black listeners gave the lowest preference rating for the Charley Pride music example. Jack Teagarden, a white performer singing in traditional blues style, received a mean score of 2.84 (§D_= 2.28) by white listeners identifying his race, and a mean score of 2.73 (§Q_= 2.21) by black listeners. The larger standard deviation shows that neither group of listeners was in total agreement when attempting to identify Teagarden's race. Nonetheless, the listeners' responses gravitated more toward 1, the "black" racial identifier than 7, the "white" identifier. The validity examples by the female artists were not easily perceived by the respondents. Mean scores and standard deviations show that the listeners were uncertain of both perfommers' race. In the answer booklet for the listening tape, a response of 4 indicated that the listener 80 was not sure of the performer's race while a response of 7 indicated that the listener was certain that the performer's race was white. A response of 1 indicated that the listener was certain that the performer's race was black. The mean scores for Rickie Lee Jones, a white performer singing in a contemporary blues style, were 4.37 (§Q_= 2.34) for the white listeners and 4.51 (§Q_= 2.44) for the black listeners. Although Jones was selected as a validity check on the listening tape for her interpretation of black vocal performance styles, listeners were more inclined to identify Jones' race as white. Mean scores for Leoyntyne Price, a black performer of art music whose music example was a popular-song style, were 4.38 (§Q_= 2.13) for the white listeners and 5.17 (§Q_= 2.15) for the black listeners. Listeners' were not in complete agreement when identifying Price's race and were not completly certain of her race, but the mean scores indicate the listeners were more inclined to identify Price's race as white, particularly the black listeners. Table 5 presents preference and perception scores by the listeners' race and the performers' race. Although the difference between the mean scores of the two racial groups' perception of the black performers' race is small, the scores demonstrate that the black subjects achieved greater accuracy and demonstrated more agreement when perceiving the white performers' race. 81 Table 5 Preference and Perceptionygcoresfby the Listeners' and Performers' Race Black Performers White Performers Preference Perception Preference Perception Subjects M SD. 1.4. 8.1.3. 1.4. 8.11 8. 511 Black 4.20 1.20 2.16 .82 2.29 .85 6.28 .56 (BF30) white 3.01 1.27 2.56 .86 2.95 1.23 5.82 1.02 (BFlOZ) Note: Preference range: 1.00 = low; 7.00 = high. Perception range: 1.00 = black performer; 7.00 = white performer. 82 Mean scores of the preference ratings for the music examples indicate that, overall, the listeners preferred the music examples by the black performers' more than the examples by the white performers. The sub-group scores reported in Tables 3 and 4 show that the black listeners preferred the music examples by the black performers while the white listeners' preference ratings are distributed more evenly between the two races. Table 5 shows that the white listeners' overall mean preference score was slightly higher for the black performers. The mean scores show that the black respondents' preferences were stronger for music-examples by same-race performers and weaker for music examples by other-race performers. But the white respondents showed more tolerance for other-race performers, though their preferences were slightly stronger for same-race perfommers. For the black listeners, the effect of race was very strong. I used the non-parametric Wilcoxon signed-rank test to determine if the differences between the respondents' perception of the black performers and the white performers was statistically significant. For each racial group of listeners, I compared the listeners' perceived race response for the black performers with the listeners' perceived race response for the white performers. For the white listeners, the test yielded z = -8.75, p_= 102, p_< .01 and for the black listeners, z -7.77, p_= 80, p_< .01. These scores 83 indicate that both the white and black listeners were able to perceive differences between vocal performances by the black artists and vocal performances by the white artists. The difference between their identification of the performers' race was statistically significant. The same procedure was conducted to assess the statistical significance of the difference between the respondents' preferences for the black performers and the white performers. The black listeners showed considerably stronger preferences for music examples by black performers, and the difference was statistically signifcant: z = -7.54, p_= 80, p_< .01. But, because the white listeners' preferences for the music examples by the black performers and the music examples by the white performers were nearly equal, the difference was not statistically significant (2 = -.99, p_= 102, p_> .01). To assess the statistical significance of the relationship between the listeners' preference responses and their perception responses, correlation procedures were conducted using the non-parametric Spearman technique. Again, I examined the sample by racial sub-groups using the following response ratings as indicated in the answer booklets: for the preference responses, a rating of 1 indicated a negative response to the music example and a rating of 7 indicated a positive response to the music example; for the perception responses, a rating of 1 84 identified the performer's race as black and a rating of 7 identified the performer's race as white. An assessment of the relationship between the listeners' preferences for the white performers and the listeners' identification of the white performers' race, produced a correlation for the white respondents that showed statistical significance: .19, p_= 102, p_< .05. Although the strength of the relationship is weak, the correlation coefficient shows that the white listeners preference ratings were stronger when they identified the performer's race as white. For black listeners, the resulting correlation was -.24, p_= 80, p_< .05 and statistically significant. The negative coefficient shows that the black listeners' preference ratings were low when they identified the performers' race as white. An evaluation of the strength of the relationship between the listeners' preferences for the black performers and the listeners' perception of the black performers' race produced different results. For white respondents, the relationship was not statistically significant. But the obtained correlation for the black respondents was -.31, p_= 80, p_< .05, and statistically significant. When the black listeners identified the performer's race as black, the listeners also provided positive preference ratings. The results demonstrate same-race preferences by the black and white listeners and show that the listeners' preferences for the music examples were related to the 85 listeners' perception of the performer's race. When white listeners identified the performers' race as white, positive preferences were provided. The black listeners' same-race preference was even greater. Moreover, the black listeners provided positive preference ratings when they identified the performers' race as black and negative preference ratings when they identified the performers' race as white. Analysis of the Rapisl-Encountsr Messurement I evaluated the subjects' responses to the attitude statements that described different racial encounters. Assessing the scale's reliability, I divided the scale into three sub-groups: responses to encounters with other blacks, responses to encounters with other whites, and responses to encounters with mixed-couples or mixed groups of blacks and whites. The obtained reliability coefficients for the sub-scales and the total scale were good: black- encounters = .95, white-encounters = .94, mixed-encounters = .91, and the coefficient for the entire scale = .91. Overall mean scores for the measurement show that combined-race listeners' gave essentially equal responses to black-encounters, white-encounters, and mixed-encounters. The possible range of scores was 1.00 to 7.00: a rating of 1.00 indicated negative response to the racial encounters, and a rating of 7.00 identified positive response to the - 86 Table 6 Mean Scores for Responses to Social Encounters with slacks and Whites Black Encounters . White Encounters Group E .82 E SJ; Combined Blacks 4.81 2.69 4.23 2.55 (nsBO) Whites 4.43 2.71 4.74 2.73 (EFIOZ) College Blacks 4.82 2.52 4.14 2.20 (QF55) White 4.69 2.36 5.03 2.50 (ns48) Middle-School Blacks 4.79 1.92 4.46 1.78 (ns24) Whites 4.20 2.50 4.48 2.49 (8554) Note: Rating for social encounters with blacks or whites: 1.00 = negative response and 7.00 = positive response. 87 racial encounters. The average mean scores for all respondents were 4.58 for black-encounters, 4.55 for whiteencounters, and 4.68 for mixed-encounters. The mean scores show slight differences in the sub- groups' responses to the racial encounters. The black respondents' mean score for the attitude statements of black-encounters was 4.81. For the same statements, the white respondents' overall mean score was 4.43. For attitude statements of white-encounters, the black respondents' mean score was 4.23 and the mean score for the white respondents was 4.74. These scores indicate that both the black and white sub-groups showed slightly greater preferences for same-race group encounters. Between the two racial groups, the average score for mixed-encounters was almost equal: 4.74 for black respondents and 4.68 for white respondents. To determine the statistical significance of the difference between the responses to the black-encounter statements and the responses to the white-encounter statements, I conducted an analysis using the Wilcoxon signed-ranks test for the two sub-groups. Because the attitude statements that addressed mixed-encounters were designed to assess responses to maxed-couples and mixed- groups of blacks and whites, I did not include the four mixed-encounters statements in this analysis. An analysis of the difference between the responses to black-encounter 88 statements and white-encounter statements by the white subjects' showed that the difference was statistically significant where s_= -3.37, p_= 102, p_< .01. For the black respondents, the difference was also statistically significant, s_= -3.97, p_= 80, p_< .01. The scores confirm the same-race preferences observed in the descriptive analysis of the racial encounter statements: the subjects responded differently to the white-encounter statements and the black-encounter statements, and the differences were statistically significant for both the black and white subjects. To assess the relationship between the respondents' racial attitudes and their preferences for the music examples by the racially different performers, I conducted a correlation analysis using the Spearman technique for each of the sub-groups. For white respondents, the relationship between the expressed preferences for black performers and responses to the black-encounter statements was not statistically significant. But the relationship between the white respondents' preferences for the white performers and their responses to the white-encounter statements was statistically significant, yielding a correlation of .18, p_= 102, p_< .05. For the black respondents, neither the relationship between their expressed preferences for the white performers and their responses to the white-encounter statements, nor their preferences for the black performers 89 and their responses to the black-encounter statements was statistically significant. The Combined Effects of Rsce spg_§chool:sevel To assess the effects of the subjects' race and school- level, I conducted descriptive analyses for both measurements to observe mean scores and standard deviations for black middle-school respondents, white middle-school respondents, black college respondents, and white college respondents. Evaluations of the newly-foamed sub-groups' responses were similar to those for the black- and white sub-groups. The first analysis was performed to determine the statistical significance of the relationship between the preference responses and the perceived-race responses. I then examined the statistical significance of differences between the listeners' ability to accurately perceive the racial identity of the white performers and the racial identity of the black performers. I also analyzed responses to the racial-encounter attitude statements to examine the relationship between the racial attitude responses and the expressed music preferences of the subjects. An examination of the listeners' perceptions of the perfommers' race shows that, overall, the black college respondents demonstrated greater accuracy when perceiving the race of the performers. At the college level, each racial group was slightly better when identifying same-race 90 perfommers. At the middle-school level, black respondents were slightly better than the white respondents when identifying the performers' race, and the black respondents' provided more accurate perceptions of the white performers' race. Average mean scores and standard deviations for the listeners' preference responses to all the black performers, white performers, and the listeners' perceptions of all the black performers and the white 91 Table 7 College Listeners Overall Preference and Perception Scores Black Performers White Performers Race Preference Perception Preference Perception White Listeners (p_= 48) M, 3.26 2.45 3.25 6.55 §__ 1.14 .84 1.08 .41 Black Listeners (p_= 56) M, 4.46 1.99 2.36 6.38 §2_ .94 .77 .88 .54 Note: Positive preference rating preference rating = 7.00. performer; perception rating of 7.00 1.00; negative Perception rating of 1.00 = black = white performer. 92 Table 8 Middle-School Listenegs' Oversll Preference and Perception Scores Black Performers White Performers Race Preference Perception Preference Perception White Listeners (p_= 54) M_ 2.79 2.65 2.69 5.17 §_. 1.35 .86 1.29 .96 Black Listeners (p_= 24) M_ 3.58 2.56 2.12 6.06 §Q_ 1.50 .81 .79 .56 Note: Positive preference response = 7.00; negative preference response = 1.00. Perception rating of 1.00 = black performer; perception rating of 7.00 = white performer. 93 performers are presented for college black and white listeners, and middle-school black and white listeners in Tables 7 and 8. Table 7 shows the scores for the college subjects by race and Table 8 shows scores for the middle- school subjects by race. The average mean scores show that the black college subjects preferred the black performers more than the white performers. The average mean scores for the white college subjects show virtually equal preferences for the black performers and the white performers. This preference tendency is similar to those of the middle-school respondents: the white middle-school listeners overall preferences were generally low, but almost equal between the black and white performers. The black middle-school respondents preferred the examples by the black performers much more than the music by the white performers. Was the difference between the listeners' identification of the black performers' race and the white performers' race statistically significant? All groups-- the white college and middle-school respondents and the black college and middle-school respondents' scores showed statistical significance beyond the .01 level of probability. The following were computed with the non- parametric Wilcoxon statistic for each sub-group: white college respondents, s_ -6.03, p_ 48, p_< .01; black -6.51, n 56, p_< .01; white college respondents, s 94 middle-school respondents, s_ -6.36, p_ 54, p_< .01; black middle-school respondents, s_ -4.29, p_ 24, p_< .01. Was the difference between preferences for the black performers and preferences for the white performers statistically significant? Using the Wilcoxon statistic, differences between the sub-groups' preference responses for black and white performers were assessed. For both school-levels, white respondents' preferences for the black and the white performers did not yield statistically significant differences. But both the black college and black middle-school listeners' preference differences for the black and the white performers yielded statistically significant differences: black middle-school respondents, 24, p_< .01; black college respondents, s_ -3.44,g; s_ -6.51, p_ 56, p_< .01. Was there a statistically significant relationship between each sub-group's perceived-race identification of the white performers' and their expressed preferences for white performers? Comparing the ranked preference and perception scores for each sub-group with the Spearman technique, no statistical significance was found for the black middle-school, white middle-school, and white college respondents. Because of the direction of scaling, the correlations between the preference and perception response scales sometimes produced negative correlations. The black college respondents yielded a negative correlation of -.37, 95 p_= 56, beyond the .05 level of significance. When the black listeners identified the performer's race as white with perception ratings near 7.00, they also demonstrated their dislike for the music example with preference ratings near 1.00. Was the relationship between each sub-group's identification of the black performers' race and each group's expressed preferences for black performers statistically significant? Again, statistical significance was shown for the black college respondents only, yielding the following correlation: -.27, p,= 56, beyond .05 level of significance. When the black listeners identified the performer's race as black with ratings near 1.00, they gave positive preference ratings near 7.00. The black respondents' same-race preferences for the music examples is clearly demonstrated with the analysis, and shows a strong preference for the music examples by performers whose race the black listeners identified as black. A descriptive analysis of responses to the attitude statements that described different racial encounters showed a distribution of somewhat equal scores across the different racial encounters for each sub-group. But some same-race preferences for racial encounters are revealed in the overall mean scores of the college listeners: the white college respondents' scores favored the white-encounter statements more than other sub-groups, and the black college 96 respondents' scores favored the black-encounter statements more than other sub-groups. The other sub-groups' scores-- black middle-school and white-middle-school--showed a slight increase in positive responses for same-race statements. Was the difference statistically significant? For each race of college subjects, the difference was statistically significant: white college subjects, s,= -3.19, p_= 48, p_< .01; black college subjects, s,= -4.57, p,= 56, p_< .01. I examined the relationship between the sub-groups' expressed music preferences and the responses to racial encounters. None of the subjects' responses to white racial encounters and preferences for white performers yielded statistically significant correlations. For responses to black racial encounters and expressed preferences for black performers, no sub-group yielded statistically significant correlations. ObservingiRespondents' Behsvior As a confirmation of the validity of the listening measurement, for each administration of the test, I observed student behavior. With my own body language, I attempted to appear unaware of student behavioral responses. The college subjects' behavior was less overt, but some music examples elicited both positive and negative responses from these subjects. For nearly all classes tested, Randy Crawford, a black female rhythm and blues vocalist, elicited positive 97 behavioral responses. But, for one of the college-level classes tested, overt behavior was difficult to ascertain. The class was predominantly white and the subjects revealed little by way of behavioral response to all of the music selections. Louis Jordan, a black male popular jazz vocalist, elicited positive responses from the middle-school students, but some college students seemed somewhat disgruntled with Louis Jordan's humor. Among middle-school listeners, blacks and Hispanics were more responsive to the example by the black folk singer Bill Broonzy. During the testing, there were many comments by the middle-school respondents regarding what they considered a lack of music with a discernable "beat," and some interrupted to ask if there were any "rap" music examples. While listening to Odetta, some gave conflicting responses, verbally expressing dislike for the example by this black female folk singer, and at the same moment, tapping their toes in response to the beat. College respondents' overt behavior was sometimes limited to smiles or other facial expressions showing like and dislike. Generally, the black respondents expressed positive responses to the examples by black performers, and even more for the gospel examples by black performers and the popular style by Randy Crawford. When displaying overt behavior, white respondents showed dislike for all of the gospel examples, but particularly the gospel examples by 98 black performers in traditional style. Overall, whites displayed more tolorant behavior for all of the music examples. Essminingiths Open-ended Responsss.‘ On completion of the listening test, I requested that the subjects provide constructed responses expressing positive and negative comments for the listening tape's music examples. The responses were classified, with similar responses placed under broad categories. The request for responses elicited both positive and negative statements from the listeners that addressed particulars of the different music styles and the performers, specific characteristics of the music, and certain aspects of the test. Some of the statements addressed the performer's race or a performer's use of ethnic-sounding speech. Of the comments with attention to "things disliked," 169 comments expressed dislike for all of the music styles and for some specific music styles. Of the styles named specifically, the country examples received 44 negative votes. The gospel examples were second to country as the most disliked style, receiving 9 negative votes. Seventy comments addressed the "old" or "outdated" sound of the music styles. Twenty-seven comments expressed dislike for the vocalists' singing. Twelve comments addressed the lack of a discernable beat and active rhythms in the music 99 examples, and eleven comments expressed dislike for the slow-tempo music examples. Regarding the specific race or ethnicity of the performers, four comments expressed dislike for the white performers, two expressed a dislike for the "British- sounding" performer (Tom Lehrer, singing in affected British style), and eight comments expressed resentment for the assumption that a specific music style should identify a specific race. Behavioral observations indicated that some listeners thought that some of the styles, for example country, made the performer's racial identification too easy. As discussed earlier, the verbal responses by Hispanic listeners indicated that they felt discriminated against because of the two measurements' focus on black and white racial groups. But the written responses of all participants revealed little reference to the racial content of the measurement. Through their behavioral responses, some of the middle-school respondents showed discomfort when responding to the racial-encounter statements. Constructed responses were not requested on completion of the racial- attitude measurement. Two of the classroom teachers, one middle-school and the other college, commented that many students may be reluctanct to disclose their personal feelings on racial matters. One teacher stated that even though the measurements' racial content elicited feelings of 100 discomfort among some of the students, the students may have experienced even more discomfort had they revealed those feelings in the constructed responses. Of the listeners' comments addressing "things liked" in the music examples, 122 expressed positive regard for the different music styles. When addressing specific styles, the gospel examples received more positive comments (19), and the jazz and blues styles received 11 and 9 comments, respectively. Other positive style-related comments were sprinkled among the remaining styles heard on the listening tape, including two positive comments for the country examples. Some of the respondents liked the "newness" of the different music examples, while others liked the familiarity of the music examples as the styles were similar to those played by parents or grandparents during their youth. Twelve comments addressed these areas. Ten comments expressed positive responses for the "funny" examples; the lyrics of music examples by Tom Lehrer and Louis Jordan were somewhat humorous. With five comments, some subjects reported that they enjoyed the challenge of trying to perceive the race of the performers. Regarding the performers' race, nine comments were positive for the use of different race performers. Specific to characteristics of the music were 99 comments. Of those comments, 9 were positive expressions for the music instruments, 27 were positive expressions for examples that were "mellow" and 101 "easy listening" and 7 were positive expressions for the slow examples. Forty responses were positive comments about examples with a discernable "beat" and the fast tempo examples and music selections that were rhythmically active. Among both the positive and negative comments, only two artists were mentioned by name with three comments for each artist: Frank Sinatra and Lou Rawls. CHAPTER V Discussion Researchers examining the effects of different listeners' characteristics, the listeners' environment, and characteristics of different music stimuli continue to affirm.Hevner's (1935; 1936) early speculations about music preference formation. One listener characteristic, race, was examined in this study. But unlike many of the earlier examinations of race and ethnicity on listeners' music preferences, I examined the effects of both the listeners' race and the performers' race. First, a determination of the listeners' ability to accurately perceive the performers' race was neccessary. Secondly, the listeners' preference for the music example was examined followed by an examination of the relationship between the listeners' expressed preference for the music examples and the listeners' perception of the performers' race. Some of the written comments from the respondents suggested that the listeners depended on the music style of the listening example when determining the performers' race, and an analysis of responses to the validity check examples indicated that many listeners' did depend on the musical 102 103 style when identifying the performers' race. Most listeners incorrectly identified Charley Pride's race as white, and the listeners' assumption was probably based on the singer's country music style performance. Although the strength of listener agreement was less, Jack Teagarden was incorrectly identified as black, and the assumption was probably based on the performer's blues style and use of black-influenced vocal stylings. For these music examples, both performers adhered to the vocal performance practices of the style in which they were singing. One of the song titles selected for the listening tape, "It Was a Very Good Year," was sung by two different performers: Lou Rawls and Frank Sinatra. Lou Rawls, a black male singer, performed the music selection in a fast tempo with a jazz band arrangement and lots of rhythmic activity and jazz vocal stylings. Frank Sinatra, a white male singer, performed the piece in a slow tempo with string instruments and a minimum amount of rhythmic activity, and his performance displayed little influence of black vocal stylings. The Sinatra performance was the first example on the listening tape and the Rawls performance, heard as example twelve, appeared later on the tape. In three of the classes tested, two middle-school and one college class, the Rawls example prompted verbal responses from a few of the listeners who stated that the example had been heard earlier and therefore played twice. Some listeners, recalling the 104 lyrics, may have assumed that the Rawls and Sinatra examples were identical and simply duplicated their earlier response to the Sinatra example for the Rawls example. The average perception score for the Sinatra example was 6.15 with a standard deviation of 1.5, while the Rawls average perception score was 3.33 with a standard deviation of 2.5. A score of 7.00 indicated that the listeners were absolutely certain the performer's race was white, while a score of 1.00 indicated that the listeners were absolutely certain the performer's race was black. These scores show that the listeners were fairly certain that Sinatra's race was white but many listeners were not exactly sure of Rawls' race, although the rating shows that, along the perception continuum, listeners' identified Rawls' performance and race closer to black than white. An analysis of individual mean scores by the listeners' race shows that the black listeners were a little more successful at discerning Lou Rawls' race than the white listeners. Table 9 shows these scores. For the black listeners, the mean score was 2.34 (SD = 2.25) and for the white listeners, 3.66 (SD = 2.46). The large standard deviations demonstrate considerable disagreement among the two racial groups of listeners for this music example. Table 9 105 Race Perception scores for Frank Sinatra and Lou Rawls Listeners Performer White Black Rawls M_ 3.66 2.34 §__ 2.46 2.25 Sinatra M_ 5.92 6.60 SD 1.67 1.15 Note: Perception rating of 1.00 = black performer; perception rating of 7.00 = white performer. 106 Average scores and standard deviations for the individual music examples show that, when listening to the taped examples, the listeners were more certain of the white performers‘ race than the black performers. For the white performers, more listeners selected a rating of 7.00 to indicate that they were absolutely certain that the performers' race was white. But when listening to the black performers, fewer listeners selected a rating of 1.00 to indicate that they were absulutely certain that the performers' race was black. The strength of the overall perception mean score for white performers, 6.01 (SD = .86), was greater than the overall perception mean score of 2.46 (SD = .88), for the black performers. Of the white performers, Rita Coolidge and Anita Bryant received the lowest average perception scores; both were female performers. Coolidge's performance was in a bluesy vocal style, and the Bryant selection was a traditional gospel piece. From the white listeners, Coolidge received an overall mean score rating of 4.48 (SD = 2.26) and Bryant's overall mean score rating was 4.73 (SD = 2.23). The listeners' decision on Coolidge's race may have been based on the blues musical style, and not on other possible indications of the performer's race. Coolidge's performance included some of the vocal stylings frequently used by black performers of the blues and rhythm and blues musical styles. But the overall performance did not include 107 some of the characteristics described in the writings of Southern (1971; 1983), a scholar of black musical performance. Bryant's performance included some instrumental rhythmic activity with a slightly syncopated melody. The musical styles of both performers probably contributed to the uncertainty indicated in the listeners' identification of the performers' race. Rita Coolidge was the only white performer receiving an average mean perception rating below 6.00 by the black listeners: mean score = 3.61 and standard deviation = 2.29. Although the black listeners' identification rating of Coolidge's race was less accurate than the white listeners, the black listeners were more certain of Bryant's race (mean score = 6.17, standard deviation = 1.65). Because many blacks hold strong affiliations with their community churches, the black listeners' may be more familiar with black gospel music performance and were probably not confused by the vocal stylings in Anita Bryant's gospel performance. Both racial groups of listeners were uncertain of Odetta's race; Odetta being a black female whose performance on the listening tape was in folk style. The folk style has traditionally been dominated and popularized by white performers. The average scores for the black and white listeners show that perhaps the combination of the folk style and Odetta's own personal style made perception of the 108 performer's racial identity difficult. For black listeners, Odetta's average perception score was 4.75 (SD = 2.58) and for white listeners, the average perception score was 4.30 (SD = 2.41). The difference, though small, may demonstrate a slightly greater tendency for black listeners to associate the folk style with white performers. Overall, when perceiving the performers' race, the black listeners were more successful than the white listeners. But, for all of the listeners, dependence on the musical style may have been crucial to identification of the performers' race. If the listeners did depend on the musical style to identify a performer's race, and if the listeners were unable to perceive differences in the two performances by Lou Rawls and Frank Sinatra, then perhaps the listeners were unable to hear and perceive subtle differences in some vocalists' performance styles. As reported, other performers elicited uncertainty among the listeners: for both racial groups of listeners, Odetta and Rita Coolidge; and for the white listeners, Anita Bryant. Although the listeners were more successful when identifying the white performers' race, their inability to perceive both subtle and overt performance characteristics of the black vocal performers suggests that some listeners may not recognize certain perfommance practices as representative of a racial or ethnic group. Table 10 shows an analysis of perception responses 109 within the different style categories indicating that certain styles were perceived by listeners as representative of one racial group more than another. Although black and white performers were equally represented in each style category, the overall mean score for the gospel performances shows that the listeners closely identified this style with black performers, while the mean scores for the folk performers shows that listeners closely identified this style with white performers. The average score for performances within the popular style category, a style more familiar to the listeners, shows that the listeners probably identified the individual performers race with fair to good accuracy. Table 10 Race Perception Scores for Different Music Styles Music Style M_ §Q_ Jazz 3.97 1.15 Gospel 3.67 .69 Folk 5.09 .99 Popular 4.20 .72 Note: Perception rating of 1.00 = black performer; perception rating of 7.00 = white performer. 110 I observed the listeners' behavioral responses to the music examples and noted that many listeners provided immediate race perception responses to the Charley Pride music example, which was a validity check for the listening tape. It appeared that the listeners did not respond to the performer, but rather the music style. The average score to the Charley Pride example, indicating that the listeners identified the performer's race as white, confirms the observed behavioral responses. In some classes, some listeners responded verbally to this example, stating that identification of the performer's race was obvious. Constructed responses provided at the end of the test confirm the listeners' dependence on the music style for help in making this judgment. Central to many comments from the constructed responses were aspects of the different musical styles. Because of the current access to video productions of popular music through the media, young listeners may not be sensitive to differences in vocal performance styles. Dependence on visual cues when identifying a performer's race may desensitize listeners to subtle differences in vocal performance practices that have been traditionally associated with members of different racial groups. Some may argue that a sort of musical "melting pot" is a noble idea that may result in music performance and listening practices void of racial prejudice. But recognizing and 111 preserving the cultural contributions of different races and ethnic groups, demonstrated in the groups' music, is equally important. When black listeners identified the performers' race as white, they frequently provided negative preference responses but when the black listeners identified the performers' race as black, they provided positive preference responses. However, the preference responses made by white listeners demonstrated greater flexibility in their preferences for the black performers. The findings in this study show that white listeners demonstrated greater flexibility in their listening preferences for different race performers. But black listeners preference for black performers was very strong. Moreover, the black listeners were able to accurately identify both the same race and different race performers and showed little flexibility in liking the white performers. With regard to styles, patterns of dislike emerge among the white listeners for the gospel performances. Among the black listeners, patterns of dislike emerge for the folk performances. For some listeners, the two styles may represent extremes of black and white performance style. Traditional black gospel style is still heard in many black churches and the church remains a strong institution within the black community. From behavioral observations of the listeners, I noted that many black listeners responded with 112 positive affect to the gospel music by the black performers. But the folk music examples elicited giggles and other negative responses from many of the black listeners. The white listeners' responses to the gospel music examples were negative and many listeners showed immediate dislike for the black performances in gospel style. I examined the effect of different age levels on the listeners' preference decisions and found that the preference decisions were similar for the middle-school listeners and the college listeners. Black middle-school and black college listeners preferred the music examples by performers they identified as black, while white middle- school and white college listeners' preferences for performers of both racial groups were equal. When perceiving both the black and white performers' race, black middle-school listeners were more successful than white middle-school listeners while black and white college listeners attained greater success when identifying same- race performers. All groups identified the white performers' race more successfully than the black performers, but the white middle-school listeners overall mean score for identifying the white performers' race was low. Overall, middle-school listeners' preference responses were lower than those of the college listeners. This finding is consistent with the earlier discussions of music 113 preference research examining the effect of listeners' age. Central to the discussion earlier in this study were reports of lower preference ratings provided by junior high and middle-school listeners, and higher preference ratings from college-age listeners to unfamiliar music examples. In this study, the listeners' race and age functioned as cultural and social influences on their music preference decisions for music by different- and same-race performers. Reporting greater "ethnic-flexibility" among whites, attitude research in the social sciences suggests that blacks' same-race preferences increase with age while whites' same-race preferences decrease with age. The black listeners' preference for black performers was considerably stronger than the white listeners' preference for white performers. Moreover, the difference between black college listeners preferences for black and white performers was greater than the preferences for different-race performers among black middle-school listeners. This finding showed that the same-race preferences for black listeners was stronger at the college level. But among the white listeners, the listener-flexibility was greater at the college level. The white college listeners showed virtually equal preferences for the white and black performers, consistent with the attitude research in the social sciences. Consistent also with the attitude research reported in Aboud and Skerry (1984), I found that within the 114 two racial groups, the strength of the listeners' responses, both positive and negative, was greater at the college level. Other findings in this study seem to suggest that group loyalty was a salient factor for not only the black respondents, but for the Hispanic minority group members as well. The Hispanic subjects, showing loyalty to their cultural group, voiced concern that their group was not represented in the study. The black listeners' preferences for the black performers demonstrated loyalty by the group members to their own cultural group. For the black listeners, the relationship between the listeners' preference for the performers and their identification of the performers' race was strong. Moreover, the strength of the relationship between the black listeners' preference for black performers and their identification of the black performers' race was greater for the black college listeners. The findings also indicate that black listener- flexibility was less at the college level. Aboud and Skerry's (1984) definition of an ethnic group as a "socially and/or psychologically defined set of people who share a common culture or cultural background" seems to point to the social and cultural cohesiveness expressed in the uniformity of responses by black and Hispanic listeners to the measurements. Aboud and Skerry (1984) report that for black, Hispanic 115 and other minority members, the motivation for own-group preferences may be different from white own-group preference. The black and Hispanic listeners' loyalty to, and expressed desire for same-race performers, may reflect the group members' need to protect their own music and other symbols representative of their groups' cultural environment. An analysis of responses to the black- encounter and white-encounter attitude statements showed some same-race preferences for both the black respondents and the white respondents. But the difference between responses to black-encounter and white-encounter statements was statistically significant for only the black college and white college respondents. This indicates that racial- flexibility was greater for middle-school respondents than for the college respondents. I examined the relationshiop between the listeners' expressed attitudes toward racial-encounters and their expressed music listening preferences. The relationship between the white listeners preference for the white performers and their preferences for social encounters with whites was strong and positive. The relationship between black listeners' music preferences and preferences for racial-encounters with blacks was not strong. 116 Conclusions_snd Recommendstions. A premise for this study was that many music educators attempt to modify the music preferences of their students. In the literature review, I reported that music preference studies have examined the effects of specific listener characteristics, the music stimulus, and the listeners' environment. In addition, I examined attitude research in the social sciences that explored the effects of peer groups, social groups, and race and ethnicity on attitude development. The results reported in this study support findings in research examining the effects of listeners' age and race on music preferences and studies of the development of ethnic attitudes among humans. Studies by Appleton (1970/1971), Meadows (1970/1971), and Killian (1990), found that the white listeners exhibited more flexibility for music examples by black and white performers. But in these studies, the black listeners listener-flexibility was less: across age groups, black listeners preferred performances by the black artists more than performances by the white artists. Although same-race preferences were reported in each of the studies cited, the same-race preference of blacks was consistently stronger. The following findings emerge from this study. The black listeners preferred music examples by black performers and the black listeners' preference for same-race performers 117 was greater among the black college listeners. Overall, the same-race preferences of black and white listeners were apparent in not only the music preference responses but also in the racial-encounter responses. Moreover, the relationship between white listeners' preference for white performers and their preference for white-encounters was strong and positive. But white listeners also demonstrated greater listener-flexibility in preferring music examples by both the black performers and the white performers. Broadening the music preferences of listeners will require that music educators become sensitive to the needs of the ethnic and racial groups represented in their classrooms. Black, Hispanic and other minority students need to know that the music contributions of their own ethnic groups are acknowledged by the educator and presented in the classroom for their own value. All students should be taught subtle and overt differences in the performance practices of blacks, Hispanics, and other minority groups as well as the cross-cultural contributions demonstrated in American music forms. Broadening students' music preferences will require that music educators discover recordings by blacks and other minorities in different music styles, and not simply focus on the styles traditionally associated with each group. Non-minority group members will also benifit from the exposure; the white students' 118 listener-flexibility will enable them to be open to music by black performers. The findings presented in this study show that black students' same-race preference requires immediate attention by the music educator. It is essential that classroom teachers include music performances by blacks when teaching black students and this need becomes even more significant when teaching older black students. If ethnic-flexibility is greater among younger children, and listener-flexibility--the listeners' willingness to explore unfamiliar music styles--is greater among younger children, then expanding music preferences must begin with the early grades. Further research in this area should include a larger sample of black middle-school respondents. An examination of other age-groups may reveal the pivotal points of listener-flexibility among whites and blacks in terms of their same-race music preferences. The study should be expanded to include Hispanics and Latins and other minority group members as listeners and possibly performers. APPENDICES SOCIPREF STUDID ETHID 1-3 6 AGE TEA‘HMR 4-5 7-8 DIRECTIONS: For each of the following statements, tell us what you would do. There are no wrong answers. With an X, mark the space somewhere between I WOULD and I WOULD NOT to show what you would do. If you can't decide, mark your X in the middle space. Be sure to mark only one X for each question. 1. If I needed to know what time it was, and there was a younger woman and an Older woman waiting at the bus stop, I would ask the younger woman. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (9) ********** 2. I would feel comfortable going to the mall with a group of black friends. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (10) ********** 3. If on the bus to ‘the :mall there *were only' two seats available, one next to a black woman and one next to a white woman, I would sit next to the white woman. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (11) ********** 119 120 4. If I needed to transfer to another bus and the only seat available was one next to a black man, I'd just stand. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (12) ********** S. If I were buying a gift for a friend, I would trust an older man to help me. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (13) ********** 6. I would feel comfortable going to the mall with a group of white friends. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (14) ********** 7. If my black friend was sick and could not go to the mall, but my white friend could, I'd just go by myself. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (15) ********** 8. If I were buying new shoes at the mall and two men, an older salesman and a younger salesman, were available to help, I'd ask the older man for help. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (16) 121 9. If I saw some white friends in the mall and they invited me to have lunch with them, I'd make up an excuse and not go. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (17) ********** 10. If I saw a black friend with her white boyfriend at the mall, I would pretend that I did not see them. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (18) ********** 11. If I lost my wallet while shopping at the mall, I would trust a black security guard to help me. I WOULD NOT | | | | | | | | I WOULD. 1 2 3 4 5 6 7 (19) ********** 12. If I were buying new shoes at the mall and two women, a black and white salesperson, were available to help me, I'd ask the black woman for help. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (20) ********** 13. If I and a group of my white friends were getting together to go to the mall, I would feel comfortable asking some of my black friends to come along, too. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (21) 122 14. If I saw a white friend with her black boyfriend at the mall, I would pretend that I did not see them. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (22) ********** 15. If I lost my wallet while shopping at the mall, I would trust a white security guard to help me. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (23) ********** 16. If I were buying new shoes at the mall and two women, a black and white salesperson, were available to help me, I'd ask the white woman for help. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (24) ********** 17. If I were buying a gift for a friend and two women, an Older woman and a younger woman were availble to help me, I would ask the younger woman. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (25) ********** 18. If I saw some black friends in the mall and they invited me to have lunch with them, I'd make up an excuse and not go. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (26) 123 19. If I and a group of my black friends were getting together to go to the mall, I would feel comfortable asking some Of my white friends to come along, too. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (27) ********** 20. If my white friend was sick and could not go to the mall, but my black friend could, I'd just go by myself. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (28) ********* 21. If on the bus to the mall there were only two seats available, one next to a black woman and one next to a white woman, I would sit next to the black woman. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (29) ********** 22. 21f I needed to transfer to another bus and the only seat available was one next to a white man, I'd just stand. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (3O) ********** 124 23. If I needed to know what time it was, and there was a younger man and an older man waiting at the bus stop, I would ask the younger man. I woum NOT | | | | | I | | I wow: 1 2 3 4 5 6 7 (31) ********** 24. If on the bus, there were only two seats available, one next to an older woman and one next to a younger woman, I would sit next to the older woman. I WOULD NOT | | | | | | | | I WOULD 1 2 3 4 5 6 7 (32) APPENDIX B nunn STUDIO: CLASSID: 1.3 ‘-s 1. SCHOOL NAME: I 2. GRADE LEVEL: 1 3. TEACHER: 4. AGE: L9 5. MALE FEMALE m 6. WHITE BLACK HISPANIC OTHER: n 7. MUSICAL INSTRUMENT? YES NO u CHOIR? YES NO B 125 126 **************$#**** PRACTICE EXAMPLE (someone else's answers) IDISLIKE I _I__I_I__I__I____I___I ILIKE BLACK I I I I..____I_____I__I___I WHITE I 2 3 4 5 6 7 YOUNGER I_I_I_I__I__I_I I OLDER 30 (use) (What do you think?) I DISLIKE I _I__I__I__I__I___I_I I LIKE I 2 3 4 5 6 7 BLACK I_I_I_I—I_I_I_I WHITE 1 2 3 4 5 6 7 YOUNGER I_I_I_I I I I I OLDER I DISLIKE 00 BLACK 0” YOUNGER I DISLIKE UH BLACK UH YOUNGER I DISLIKE 00 BLACK 0” YOUNGER **********¢****#**** I LIKE WHITE OLDER I LIKE WHITE OLDER I LIKE WHITE OLDER I DISLIKE UM BLACK 0” YOUNGER I DISLIKE 0U BLACK an YOUNGER I DISLIKE an BLACK an YOUNGER 30 (use) **********#********* WHITE OLDER I LIKE WHITE OLDER I LIKE WHITE OLDER I DISLIKE an BLACK UH YOUNGER I DISLIKE an BLACK an YOUNGER I DISLIKE (w) BLACK U0 YOUNGER ******¢*¢¢**¥******* WHITE OLDER I LIKE WHITE OLDER I LIKE WHITE OLDER IO. 11. 12. I DISLIKE on BLACK (3) YOUNGER I DISLIKE (N) BLACK 09 YOUNGER I DISLIKE (3) BLACK an YOUNGER (use) *ttt**********t#*#** ******************** I LIKE WHITE OLDER I LIKE WHITE OLDER I LIKE WHITE OLDER 13. I4. 15. I DISLIKE (3) BLACK (3) YOUNGER I DISLIKE (w) BLACK UH YOUNGER I DISLIKE UN BLACK an YOUNGER (use) *********#***#****** WHITE OLDER I LIKE WHITE OLDER I LIKE WHITE OLDER 16. 17. 18. I DISLIKE (M) BLACK UM YOUNGER I DISLIKE UH BLACK 0” YOUNGER I DISLIKE 00 BLACK 0” YOUNGER 1 2 3 4 5 6 7 (use) ¥*¥************t#*** I LIKE WHITE OLDER I LIKE WHITE OLDER I LIKE WHITE OLDER 19. 20. I DISLIKE (N) BLACK 60 YOUNGER I DISLIKE on BLACK 6” YOUNGER I LIKE WHITE OLDER I LIKE WHITE OLDER 134 What are some of the things you LIKED about the music? What are some of the things you DISLIKED about the music? APPENDIX C To the Student: We will caiduct a research project and will need your written consent for you to participate. As a participant in the study, you will listen to sale misic exaIples and tell us what you think about the nusic. A response booklet will be used a: which you can record responses indicating if you like the music and your guess of the performer's age and racial identificatim. In the second part of the project, you will respmd to written statemmts about the kinds of people you would prefer to socialize or interact with. Again, you will respond to statements in the answer booklet by indicating what you would do in certain social settings involving people of different age groups and racial groups. We will mly ask that you give us your age, sex and ethnic identificatim and will assign identificatim numbers In the booklet provided. We do not want to know your name and it will not appear on the response booklet. All informtim will be taken cmfidentially and the results of the study will in no way identify you. (lice caIpleted, results of the study will be nude available upm your written request. The study will involve two separate class periods to be determined by your teacher. In both cases, the anount of time needed for testing will be less than thirty minutes. Your participation in the study is voluntary and is not required. You my choose not to participate at any time prior to and during testing without penalty. The project will involve two separate class periods which will be determined by your classroan teacher. If you should have any questions about this project, please contact Jan Mccrary at 347-0429. ******************** I agree to partipate in the research project described above. (Pi-Immune) (Ya: dam) 135 136 To the Parent: We will conduct a research project that will involve mariners of your child's classroan. For your child to participate in the project, we need your written cmsent. You can do this by signing the bottan of this form and having your child return it to his or her classroan teacher. As a participant in the study, your child will listai to some nusic exaIIples and tell us what he or she thinks'about the nusic. An answer booklet will be used on which your child will write his or her answers by sinply indicating if he or she likes the mic. We will also ask your child to guess the performer's age and racial identificatim. In the second part of the project, your child will respond to written statmts about the kinds Of people he or she prefers to socialize and interact with. Again, your child will write sinple respmses to the stataImts in a booklet telling us what they would do in social settings with people Of diffeer age groups and racial grows. We will ask your child to write his or her age, sex and ethnic identification m the answer booklet and we will write an identification umber on the booklet. But we do not want to know his or her name for this study, and it will not appear on the answer booklet. All information will be taken confidentially and the results of the study will in no way identify your child. mce carpleted, results of the study will be Wade available upon your written request. The project will involve two separate class periods which will be determined by the classroan teacher. In both cases, the mount of time needed for testing will be less than thirty minutes. Your child's participatim is voluntary and he or she will not be required to participate. He or she may choose not to participate at any time prior to and during testing without penalty. If you should have any questions about this project, please contact Jan McCrary at 347-0429 . ******************** has my permission to participate in (child's name) the research project described above. (parent's signature REFERENCES Referances Abeles, H. E., Hoffer, C., & Klotman, R. (1984). Foundations of magic education. New York: Schirmer Books. Aboud, P. E., Skerry, S. A. (1984). The development Of ethnic attitudes: A critical review. Journal of Croas-Cultural Paychology, 1;, 3-34. Ajzen, I., & Pishbein, M. (1977). Attitude-behavior relations: A theoretical analysis and review of empirical research. Psychological Bulletia, 81, 888-918. Allport, G. W. (1935). Attitudes. In C.A. Murchison (Ed.), A__handbook of agcial psychology (pp. 798 -844). New York: Russell and Russell. Cited in Rajecki, D. W., (1982). Attitudes. Sunderland, MA: Sinauer Associates. Alpert, J. (1982). The effect of disk jockey, peer and music teacher approval of music on music selection and preference. Journal of Reagarch in Muaig_ Education, a9, 173-186. Appleton, C. (1971). The comparative preferential response Of black and white college students to black and white folk and popular musical styles. (Doctoral dissertation, New York University, 1970). Dissertation Abstracta InternationaL,_;;, 2723A. Bandura, A. (1971). Psychological modeling, Chicago: Adline-Atherton. Bandura, A., & Walters, R. (1963). Social 1earniag_ and pergonality development. New York: Holt, Rinehart & Winston. Bartlett, D. L. (1973). Effect of repeated listenings on structural discrimination and affective response. Journal of Research in Muaic‘gducation. 2;, 302-317. Baumann, V. H. (1960). Teen-age music preferences. Journal of Research in Music Education, a, 75-84. Bussey, K., & Bandura, A. (1984). Influence Of gender constancy and social power on sex-linked modeling. 137 138 Journal of Personality and Social Psycholqu, 11, l.292-1,302. Choate, R. A. (1968). ngamentary raport of tha Tanglewood symposium. Washington: Music Educators National Conference. Cook, 8., & Selltiz, C. (1964). A multiple-indicator approach to attitude measurement. Psychological Bulletin, a;, 36-55. Darrow, A., Haack, P., 8 Kuribayashi, F. (1987). Descriptors and preferences for Eastern and Western musics by Japanese and American nonmusic majors. Journal Of Research in Muaic Education, §§J 237-248. Doyle, A. B., Beaudet, J., 8 Aboud, F. (1988). Developmental patterns in the flexibility of children's ethnic attitudes. Journal Of Croaa;_ Cultural Payghology, 12, 3-18. Edwards, A. (1957). Techniquaa Of:attitude scale conatruction. New York: Appleton-Century-Crofts, Inc. Edwards, A., 8 Kilpatrick, F. P. (1948a). Scale analysis and the measurement Of social attitudes. Psychometrika, 1;, Edwards, A. & Kilpatrick, F. P. (1948b). A technique for the construction Of attitude scales. Journal of Applied Paychology, 3;, 374-384. Farnsworth, P. R. (1950). apaical taste: Its measurement and cultural nature. Stanford: Stanford University Press. Farnsworth, P. R. (1969). The apcial psychology of music. Ames, IA: The Iowa State University Press. Festinger, L., Schachter, S., 8 Back, K. (1950). Social preaaurea‘in informal groups. Stanford: Stanford University Press. Finnas, L. (1989). How can musical preferences be modified? Bulletin for tha Council for Research in Muaic Education, 102, 1-58. Flowers, P. (1988). The effects Of teaching and learning experiences, tempo, and mode on 139 undergraduates' and children's symphonic music preferences. Journal of Research in Music Edugation, 36, 19-34. Furman, C. E., 8 Duke, R. A. (1988). Effect of majority consensus on preferences for recorded orchestral and popular music. Journal of Research in Magic Eduaation, aa, 220-231. Geringer, J. M. (1982). Verbal and operant music listening preferences in relationship to age and musical training. Paychology,of Music, [Special Issue]. 47-50. Geringer, J. M., 8 Madsen, C. K. (1987). Pitch and tempo preferences in recorded popular music. In C. K. Madsen and C. A. Pricket, Applications pf research in muaic behavior (pp. 204-212). Tuscaloosa, AL: The University Of Alabama Press. Greer, R. D., Dorow, L. C., Randall, A. (1974). Music listening preferences Of elementary school children. Journal of Research in Maaic Education. ZZJ 284-291. Greer, R. D., Dorow, L. C., Wachhaus, G., 8 White, E. R. (1973). Adult approval and students' music selection behavior. Journal Of Research in Music Education, 2;, 345-354. Hedden, S. K. (1981). Music listening skills and music listening preferences. Bulletin for tha_ Council for Research in Muaic Education. §§J 16- 26. Hevner, K. (1935). Expression in music: A discussion Of experimental studies and theories. Psychological Review, 5;, 186-204. Hevner, K. (1936). Experimental studies of the elements Of expression in music. The Americap_ Journal Of Paycholagy, aa, 246-268. Inglefield, N. G. (1972). Conformity behavior reflected in the musical preferences of adolescents. Contributions to Maaicagduaation, 1, 56-67. 140 Killian, J. N. (1990). Effect Of model characteristics on musical preference of junior high students. Journal Of Research in Muaig Education, 33, 115-123. Killian, J. N., 8 Kostka, M. (in press). An investigation of children's preferences for peer- approved music. Update: Applicationa of Research in Muaic Education. Kuhn, T. L. (1980). Instrumentation for the measurement of music attitudes. Contributiona to music education, 3, 2-38. LeBlanc, A. (1979, April). Generic style music preferences of fifth grade students. Paper presented at the meeting of Music Educators National Conference, Indianapolis, IN. LeBlanc, A. (1979). Generic style music preferences of fifth grade students. Journal Of Research in Music Education, 31, 255-270. LeBlanc, A. (1982). An interactive theory Of music preference. Journal Of Muaic Therapy, 33, 28-45. LeBlanc, A. (1984). Selecting a response mode in music preference research. Contributiona to Music Education, 33, 1-14. LeBlanc, A., Colman, J., McCrary, J., Sherrill, C., 6 Malin, S. (1988). Tempo preferences of different age music listeners. Journal Of Reaaarch in Mu§;g_ Education, 33, 156-168. LeBlanc, A., 8 Cote, R. (1983). Effects of tempo and performing medium on children's music preference. Journal Of Research in Muaic Educatioa, 33, 57-66. LeBlanc, A., & McCrary, J. (1983). Effect of tempo on Children's music preference. Journal of Research in Music Education, 33, 283-294. LeBlanc, A., 8 Sherrill, C. (1986). Effect of vocal vibrato and performer's sex on children's music preference. Journal Of Reaaarch in Music Education, 33, 222-237. Mark, M. (1986). Contemporary music education. New York: Schirmer Books. 141 Madsen, C. K. 8 Madsen, C. H. (1975). Selection of music listening or candy as a function of contingent versus noncontingent reinforcement and scale singing. In C. K. Madsen, R. D. Greer, 8 C. H. Madsen (Eds.), Research in musicttehavior (pp. 89-96). New York: Teachers College Press. Meadows, E. (1971). The relationship of music preference to certain cultural determiners. (Doctoral dissertation, Michigan State University, 1970). Dissegtation Abstracts International, 3;, 6100A. Nettl, B. (1976). Folk and traditional music Of tha_ gagtern continentg, Englewood Cliffs, NJ: Prentice-Hall, Inc. Price, R. E. (1986). A proposed glossary for use in affective response literature in music. Journal Of Research in Music Education, 33, 151-159. Prince, W. F. (1972a). A paradigm for research on music listening. Journal of Research in Music Education. 29, 445-455. Prince, W. F. (1972b). Some aspects of liking responses Of junior high school students for art music. Contrttutiong to Magic Education, 3, 25- 35. Radocy, R. E. (1975). A naive minority of one and deliberate majority mismatches Of tonal stimuli. Journal Of Research in Mugic Education. ZéJ 120- 133. Radocy, R., 8 Boyle, J. D. (1979). Psychological foundationg of mugical behavior. Springfield, IL: Charles Thomas, Publishers. Schuessler, K. P. (1980). Musical tastetand socio- aconomic background. New York: Arno Press. Shaw, M. E. 8 Wright, J. M. (1967). Scales for the measuremgnt Ofaattitudes. New York: McGraw Hill Book CO. Shehan, P. K. (1982). Student preferences for ethnic music styles. Contributions tO Mugic Education, 2: 21-28 0 142 Shehan, P. K. (1985). Transfer Of preference from taught to untaught pieces of non-western music genres. Journal Of Research in Magic Education, 1;: 149’158 . Sims, W. L. (1986). The effect of high versus low teacher affect and passive versus active student activity during music listening on preschool children's attention, piece preference, time spent listening, and piece recognition. Journal of Research in Magic Education, 33, 173-191. Sims, W. (1987). Effect of tempo on music preference of preschool through fourth-grade children. In C. K. Madsen 8 C. A. Prickett (Eds.), Applicationg ct research in music behavior (pp. 15-25). Tuscaloosa: The University Of Alabama Press. Southern, E. (1971). Music Of black Americans. New York: W. W. Norton. Southern, E. (Ed.) (1983). Readings in black American music. New York: W.W. Norton. Summers, G. F. (1970). Introduction. In G. P. Summers (Ed.) Attitude Measurement (pp. 1-17). Chicago: Rand McNally. Wapnick, J. (1976). A review Of research on attitude and preference. Bulletin for the Council for Research in MugictEducation, 33, 1-20. Webster, P. R., 6 Hamilton, R. A. (1981/1982). Effects of peer influence, rhythmic quality, and violin timbre on the musical preferences of fourth, fifth, and sixth grade children. Contpibutipns to Magic Education, 3, 10-20.