Investigation of musical taste, preference, and attitude has continued as a major research area in music education since the 1930's. Early studies concerned variables such as intelligence, aptitude, sex, age, and personality. A thorough, recent review of such variables has been presented by Wapnick.1
Many of the variables related to musical taste have been investigated by Farnsworth.2 In a number of studies, Farnsworth also reviewed the relationships among various techniques of measurement of musical taste. Measures have included eminence rankings of composers by musically trained and naive respondents, space allocations in music dictionaries and encyclopedias, orchestral program frequency, number of recordings listed in Records in Review, and frequency of radio broadcasts. Farnsworth concluded that musical taste was peculiar to a particular group of people, place, and period in history. Stability of taste was described as specific to a limited time and relatively narrow cultural area. For example, members of the American Musicological Society in 1944 agreed among themselves extremely well (rho = .97).
Several studies have indicated that musical preference is related to training. In a study of high school students, Birch found that students with at least three years of high school music classes owned significantly more serious music recordings than students with less than three years of classes.3 Kelly reported that students who studied musical instruments showed a preference for classical over other types of music.4 In a study that manipulated musical information as an independent variable, Larson found the greatest gain in preference to occur for the group receiving specific musical information.5
Among lay groups, Farnsworth found a steady increase in the degree of agreement as the groups became more alike in age and training.6 Compared to an American Musicological Society survey, fifth-graders agreed only slightly (rho = .33), sixth-graders were closer (rho of .52); the concurrence was more marked with high school students (.68) and college sophomores (.79). Farnsworth has also found agreement in taste when comparing dependent measures. College students in free recall versus recognition comparisons revealed a rho of .85.
Research has indicated that verbal reports should not be used as the only indication of media consumption of music or to predict reinforcement value.7
The present survey attempted to investigate aspects of musical taste among musically trained secondary school students, and trained and untrained college students. The study was designed to assess agreement among these three populations, as well as to compare concordance within each sample concerning frequency of mention, ranking, concert attendance, and number of recordings owned of subjects' ten favorite composers from any musical style and period.
Three hundred ninety-eight students from two college communities in Texas (Waco and Austin) responded to a free-recall survey of musical taste. Respondents were randomly selected junior and senior high school students enrolled in school music performance classes (116), college education majors enrolled in music fundamentals or music methods for the prospective elementary or special education teacher (115), and music majors enrolled in undergraduate and graduate music courses. The college music major group was selected from a list of course offerings at both the University of Texas and Baylor University. One hundred thirty-one of these surveys were returned with usable information.
Surveys were administered by the regular music teacher in all of the classrooms. The teachers gave brief instructions in addition to those on the form. These were that "composers may be from any style and period," and "students may list performers (or groups) only if they cannot think of ten composers."
The survey asked students to: a) list ten favorite composers; b) rank each in order of preference; c) write the number of live concerts attended in the past year where at least one of the composers' pieces was performed; and d) indicate the number of recordings (LP/tape) owned of each of the composers' works. Students were able to complete the form within a ten-minute period.
Table 1 indicates the frequency with which composers were reported in the ten favorites for each of the three samples.
FREQUENCY OF MENTION
|Secondary School (n = 116)||Non-Music (n = 115)||Music Majors (n = 131)|
|25||M. Ferguson||29||B. Manilow||55||Chopin|
|17||G. Gershwin||22||B. Bacharach||27||Schumann|
|17||B. Streisand||19||J. Mitchell||24||Tchaikovsky|
|15||P. Williams||16||D. Fogelberg||22||Handel|
|14||S. Wonder||12||J. Denver||17||Bartok|
|13||D. Fiske||12||W. Nelson||16||G. Gershwin|
|11||Fleetwood Mac||10||Fleetwood Mac||14||Ravel|
|10||E. John||10||L. Ronstadt||11||Liszt|
|10||C. Mangione||10||S. Wonder||11||Verdi|
Review of the Table reveals that college music students indicated marked preferences for composers in the formal tradition (96%), while the younger and non-music students preferred current popular composers (79%). Among music majors, the first ten composers accounted for 72 per cent of the votes for the twenty-five listed. The corresponding percentages for the non-music and secondary groups were less, 65 and 62 per cent.
Tables 2, 3, and 4 present the top ten composers based on a sum of ranks of the data gathered in the categories labeled frequency, average rank, concerts (attended during the past year), and records (number of LP/ tapes owned). According to Kendall, the best estimate of the "true ranking" is provided by the order of the sums of ranks.8 This is the order given in the column labeled "overall rank". Each of the numbers in the Tables lists a composers' relative standing among the complete within-sample listing.
Table 2 provides a comparison among the responding music majors for the four categories of taste. The Kendall Coefficient of concordance indicates a high degree of agreement, W = .802. The null hypothesis that the four sets of rankings are unrelated may be rejected at the .001 level of significance (X2 = 32.1, df = 10). For this sample of music majors, there appears to be a significant relationship among the number of records owned, concerts attended, average rank-order, and frequency of mention for the ten favorites.
WITHIN-SAMPLE RANKING OF TOP TEN
Table 3 also provides a within-sample comparison. Among the non-music major college students, only five composers in the formal tradition are listed. The Kendall Coefficient for these sets of ranks does not reveal a high degree of concordance, W = .34. For these non-music majors, there appears to be no significant relationship among the four rank orders.
WITHIN-SAMPLE RANKING OF TOP TEN
Similarly, Table 4 represents the overall within-sample comparison for the secondary school students. This list contains only four traditional composers, and places a popular composer first in three of the four rankings. The Kendall W for junior and senior high school students is .28. Thus it appears that the four sets of rankings are not significantly related.
WITHIN-SAMPLE RANKING OF TOP TEN
SECONDARY SCHOOL STUDENTS
Table 5 presents a comparison of the 1964 survey of the American Musicological Society9 with the present sample of college music majors. The music students' list is ordered according to frequency of mention, as in the AMS survey. It should be noted, however, that the AMS list of ten all-time eminent composers was made with reference to a provided listing of composers, while the music students recalled their favorite composers. The Spearman rank correlation coefficient was used to compare these lists, and indicated a rho of .764. This relationship was significant at the .02 level (t = 3.35, df = 8). Thus, the 1964 AMS survey and the current college music major sample appear to be significantly associated.
RANK ORDER COMPARISON OF MUSIC MAJORS WITH THE
1964 SURVEY OF THE AMERICAN MUSICOLOGICAL SOCIETY
|Composer||AMS (n = 853)||Music Majors (n = 131)|
Results of the survey indicated that college music students' favorite composers correspond to the formal tradition. Record collections and concert attendance showed a high degree of concurrence with ranking and frequency of mention. Some of this agreement may be due to greater environmental opportunities for music majors to hear concerts and records consistent with their educational program. Non-music college students and secondary school music students indicated greater preferences for current popular composers. Record collections, concerts, rankings, and frequency data did not show a significant association.
The greater variability in the secondary and non-music respondents' may be due to several factors. The popular composer favorites (particularly among the public school students) were those who had current or recent best-selling albums, whereas those composers without a recent best-seller were ranked lower. Musical taste may be stable for a shorter period of time for these populations. An exception to this was the Beatles, also identified as Lennon/McCartney. Another conceivable factor was the possibility of one or more recent public appearances in the Waco or Austin area by some of the popular composers. Although such sampling errors may limit the present generalizations, the results do conform to surveys of popular music taken during the year of 1977 and early 1978.
The comparison of college music students with the 1964 survey of musicologists showed a significant agreement. This finding supports those of Farnsworth, that taste is relative to a narrow cultural area. It does not seem surprising that increased training in music leads to increased preference for composers in the formal tradition, as well as greater homogeneity of preference. A greater percentage of the total responses was distributed among the top ten composers for the music majors than for the other two groups.
It should be remembered that collective presentation of taste data may obscure individual differences of considerable magnitude. Additionally, persons untrained in music when indicating a preference for Beethoven or Bach may have few reasons for placing them so high on a list of favorites. Such persons may not answer with knowledge, but in accordance with expectation or prestige value. The music educator concerned with assessing individual taste and preference might consider techniques to reduce such verbal-report versus behavior discrepancies. Further information regarding music behavior could be achieved by systematic observation in performance and functional contexts, such as the individual's actual concert attendance, or counting the number of recordings owned by individuals.
1J. Wapnick, "A Review of Research on Attitude and Preference," Bulletin of the Council for Research in Music Education, Vol. 48 (1976), pp. 1-20.
2R. Farnsworth, Musical Taste: Its Measurement and Cultural Nature (Stanford: Stanford University Press, 1950); "The Phenomenon of Musical Taste," Psychology Today, Vol. 1 (1967), pp. 14-19, 40-43; The Social Psychology of Music (Ames: Iowa State University Press, 1969).
3T. Birch, "Musical Taste as Indicated by Records Owned by Students with Varying High School Experience," Missouri Journal of Research in Music Education, Vol. 1 (1962), pp. 53-54.
4D.F. Kelly, "A Study of the Musical Preferences of a Select Group of Adolescents," Journal of Research in Music Education, Vol. 9 (1961), pp. 118-124.
5P. Larson, "The Effect of Musical and Extramusical Information upon Musical Preference," Journal of Research in Music Education, Vol. 19 (1971), pp. 350-354.
6Farnsworth, 1950, p. 24.
7J.M. Geringer, "An Assessment of Children's Musical Instrument Preferences," Journal of Music Therapy, Vol. 14 (1977), pp. 172-179.
8M.G. Kendall, Rank Correlation Methods (London: Griffin Press, 1948), p. 87.
9Farnsworth, 1969, p. 228.