Effects of Popular and Classical Background Music on the Math Test Scores of Undergraduate Students

Mike Manthei

Minneapolis, MN

Steve N. Kelly

University of Nebraska at Omaha

Abstract

The purpose of this study was to investigate the effects of five popular and classical background music listening styles on undergraduate students' math test scores. Students (n=72) from a required university music appreciation class were exposed to three different listening situations over an established period of time while completing three parallel forms of a math placement test consisting of 16 questions. Students also completed a questionnaire seeking to determine their type of response to music. The students, mostly from rural communities, represented a cross section of the university community and none were music majors. Regression analysis found that the music had no statistically significant effect on the math test scores. This was further supported by a post-hoc questionnaire. This study lends support to previous research indicating that background music had no effect on performance in other academic learning areas.

Effects of Popular and Classical Background Music on the Math Test Scores of Undergraduate Students

With the advent of electronically reproduced music, background music has become increasingly prevalent in our society. The occurrence of such music is so common that an individual may not be aware of music in their immediate environment. Background music can be defined as any music played while the listener's attention is focused primarily on a task or activity other than listening to the music (Radocy & Boyle, 1988). The function of background music varies with the individual listener and with the nature of the task or activity in which the listener is involved. Such a task or activity could be studying or other academic preparation. Students of all ages have often claimed that they can study and learn more effectively while listening to music. Indeed, some researchers have explored the possible transfer of cognitive abilities to other curricular areas by theorizing that exposure to music, through participation and formal instruction, can facilitate nonmusical learning (Madsen, 1987; Radocy & Boyle, 1988; Wolfe, 1983). Yet a solid research base for these claims seems to be lacking. While music appears to enhance some individuals' learning, it may be distracting to others.

The possible effects of exposure to music and music instruction on nonmusical learning have received some previous attention from the music education research community. Two reviews of literature (Hanshumaker, 1980; Wolff, 1977) discussed the extramusical effects of music education . The results of these reviews demonstrated mixed results. While some research indicated there may be some measurable effects of music instruction on cognitive skill, due to some inadequacies in the experimental designs such positive conclusions should be viewed with caution. Other researchers have utilized music as a reinforcer or mnemonic device (Madsen & Forsythe, 1973; Traver-Holder, 1993). Also, studies have attempted to draw a cause and effect relationship between music study and academic achievement (Friedman, 1959). Greenberg and Fisher (1971) discovered that background music had a statistically significant effect on psychological test scores. However, the direct effects of exposure to musical sounds during study or academic testing have received comparatively little attention.

Henderson, Crews, and Barlow (1945) found that popular music distracted subjects on a paragraph comprehension test, while it had no effect on vocabulary test scores. In a similar study, LaBach (1960) found background music had no effect on reading comprehension scores. He also discovered that the subjects' preference for listening to music while studying had no significant effect when used as a covariate. Etaugh and Michaels (1975) found an interaction between gender and frequency of studying to music that affected reading comprehension scores. However, Kelly (1993) did not find this interaction in a later study.

Studies that have examined the exposure to musical sounds on math skills have had similar results. Wolf and Weiner (1972) reported a statistically significant difference between music and silent conditions on arithmetic test scores. However, they attributed this difference to habituation as most of the test subjects reported that they listened to "hard rock" music when they studied. Wolfe (1983) found no difference in math test scores with four levels of music loudness, but did find that the subjects' reported that the "louder" music interfered with their concentration. In her dissertation, Cox (1981) reported that classical music used during relaxation therapy had no statistically significant effect on algebra scores. Moller (1980) found no significant difference in math test scores among groups exposed to three conditions: no sound, white noise background, and background music (John Cage's Fontana Mix).

Researchers have reported that the results of this body of literature reveal mixed results (Etaugh & Michaels, 1975; Madsen, 1987). However, finding optimum academic study and testing conditions for a variety of students is of interest to educators in all fields. Also, the effects of environmental conditions on learning and performance may reveal keys to the inner workings of the human thought process.

In his dissertation, Hedden (1971) proposed five music reaction profiles or music listening styles. These five styles are: associative, cognitive, physical, involvement, and enjoyment. Hedden hypothesized that people listen to music in a combination of five styles where all styles are present ar some level, but one style is predominant. How a person listens to music may affect the possible transfer of cognitive abilities to other curricular areas. It is the purpose of this study to investigate the effects of popular and classical music listening styles on undergraduate students' math test scores. Specifically, this study will seek to determine if any of the five musical styles, as defined by Hedden (1971), will act as a covariate, along with the presence of popular or classical background music, in affecting undergraduates' math test scores.

Procedures

This study utilized undergraduate students (n=72) from a required music appreciation class in a small university located in the North Central region of the country. This group of subjects was selected based on their availability and their academic and socioeconomic diversity. The subjects were primarily traditional university age students (18-22) and approximately equal in gender representation. Though some of the students may have had previous musical backgrounds, none of the students serving as subjects were music majors. The students were mostly from rural communities and represented a cross section of the university community. There was some ethnic and social diversity, however, the class could be described as homogeneous.

The independent variables were the background music conditions: classical, popular, and none; and the five music listening styles as defined by Hedden (1971): associative, cognitive, physical, involvement, and enjoyment. Hedden used all five music listening styles in testing the reaction to both classical and popular music. The dependent variables for the present study were math test scores.

Since the terms classical and popular music may cover a wide range of stylistic variance, this study used specific musical examples. Stewart (1984)found a difference between subjects' responses to verbal and operant preferences that varied among musical styles. The present study controlled for that difference by specifically defining classical and popular music with operant musical examples which the subjects were (for the purposes of this study) to consider typical of that musical style.

To begin this study, the subjects listened to a short except (one minute of the first movement) from Mozart's Divertimento No. 12 in E-flat, K. 252, performed by The New York Philomusica (CDX 5051). The proctor turned off the music and instructed the students that, for the purposes of this study, they were to consider the example typical of classical music. The proctor then asked the students to fill out Hedden's music listening reaction scale for classical music. The Hedden music listening reaction scale consists of twenty questions, each followed bv a bar scale allowing the subject to indicate the strength of this response in a scale of one to one hundred. Next, the subjects listened to a short excerpt (one minute) of ZZ Top's Somebody Else Been Shakin' Your Tree that, for the purposes of this study, they were to consider to be typical of popular music. The proctor then asked them to fill out Hedden's music listening reaction scale for popular music; also with the music turned off. The questions for the popular music listening reaction scale are identical to the classical music listening reaction scale. In his dissertation Hedden provides a formula for computing a value for the five music listening styles on a scale from one to one hundred based on the raw data.

During the next three class sessions (the class met twice weekly) the subjects completed three parallel forms of a math test, consisting of 16 questions each. The three test forms were derived from a math placement test used during the 1980's by the university's math department. The reliabilities of the three test forms for this sample were computed by Cronbach's alpha and Cochran's Q and were found to be acceptable (see Table 1).

Table 1

Reliability Estimates for the Three Math Tests

CLASTST (classical music)
N OF CASES = 68.0 N OF ITEMS = 16
Cochran's Q = 138.8470 p <.0001
ALPHA= 0.6775  
POPTST (popular music)
N OF CASES = 64.0 N OF ITEMS 16
Cochran's Q = 140.7540 p <.0001
ALPHA=0.6969  
CONTST (no background music)
N OF CASES = 72.0 N OF ITEMS 16
Cochran's Q = 132.7166 p <.0001
ALPHA = 0.6886

Before passing out math test form one, the proctor instructed the class that they would have 10 minutes to complete the 16 math questions. Mozart's Divertimento No. 12 in E-flat, K. 252 was playing at a comfortable listening level for taking the test (as determined by class consensus) and the subjects were given the choice of where to sit in a large classroom with the loudspeakers at the front. The proctor then distributed the math test during which the music continued throughout the administration of the test. After ten minutes the proctor told the students to put down their pencils and pass in the tests. The music played during the entire test administration of about fourteen minutes. During the next class session, the proctor followed the same procedure for math test form two except three ZZ Top recordings (Somebodv Else Been Shakin' Your Tree, Brown Sugar, and Squank) provided the background music. Again, the music played during the entire test administration. For the third math test form the proctor played no music during the testing.

Regression analyses were used to analyze the data. The independent variables were the music listening style scores and the math test condition with no background music (CONTST). The dependent variables were the math test scores under the music listening conditions of classical (CLASTST) and popular (POPTST). Also, as an independent variable in a post hoc procedure, students were asked to answer yes or no to the following statments: "When I study, listening to classical music interferes with my concentration" and "When I study, listening to popular music interferes with my concentration".

Results

To determine the effects of music listening style scores and the background music conditions, CLASTST (the test taken with classical background music) and POPTST (the test taken with popular background music) were regressed on their respective music listening style scores and CONTST (the control test with no background music). Table 2 lists the variable labels.

Table 2

Variable Names

CLASSICAL MUSIC LISTENING STYLES
CA Associative
CI Involvement
CE Enjoyment
CP Physical
CC Cognitive
POPULAR MUSIC LISTENING STYLES
PA Associative
PI Involvement
PE Enjoyment
PP Physical
PC Cognitive
YC When I study, listening to classical music interferes with my concentration. 1 = yes, 0 = no
YP When I study, listening to popular music interferes with my concentration. 1 = yes, 0 = no
CLASTST The math test score with classical background music
POPTST The math test score with popular background music
CONTST The math test score with no background music

Tables 3 and 4 show the results of two regression analyses. The results indicated there are no music listening style variables that have any statistically significant effect on the test score difference for either the classical or the popular music listening condition. CONTST has a statistically significant effect on the two dependent variables. This may indicate that people who perform w ell in math also perform well studying math under background music conditions.

Table 3

Regression Analysis for CLASTST Regressed on the Music Listening Style Scores for Classical Music and CONTST

  DR Sum of Squares Mean Square
Regression 6 152.75845 25.45974
Residual 42 123.48645 2.94015
R Square = .55298   F = 8.65932 Sig F<.0001
Variable B SE B Beta t Sig. t
CONTST .694126 .108881 .715638 6.375 .0000
CI -.038022 .022374 -.199713 -1.699 .0966
CC -.007180 .013090 -.077043 -.549 .5862
CP -.002731 .023938 -.018487 -.114 .9097
CE -.008017 .019952 -.077545 -.402 .6899
CA .028530 .023452 .237064 1.217 .2306
(Constant) 4.777199 2.009605    

Table 4

Regression Analysis for POPTST Regressed on the Music Listening Siyle Scores for Popular Music and CONTST

  DF Sum of Squares Mean Square
Regression 6 136.7454 122.79090
Residual 42 163.49948 3.89284
R Square = .45545   F=5.85456 Sig. F = .0002
Variable B SE B Beta t Sig. t
PE .016287 .024909 .146682 .654 .5168
Pi -.001401 .019273 -.008721 -.073 .9424
CONTST .699301 .127056 .691558 5.504 .0000
PC .004390 .016337 .036949 .269 .7895
pp -.026498 .03658.4 -.138622 -.724 .4729
PA -.004486 .027663 -.032442 -.162 .8719
(Constant) 6.972756 4.639115      

A t-test was used to analyze the differences among the test score means for CLASTST, POPTST, and CONTST. Table 5 presents the comparison of math test score means and their corresponding t-tests. The t-test revealed no statistically significant differences among the three testing conditions. One may assume that the presence of either classical or popular background music had no effect on the subjects' math test performance.

Table 5

Comparison of Math Test Score Means

Variable Number of Cases   Standard Deviation Standard Error
CLASTST classical) 62 11.8387 2.681 .340
POPTST (popular) 62 12.0806 2.675 .340
t = -.93 p=.354        
Variable Number of Cases   Standard Deviation Standard Error
CLASTST (classical) 66 11.8485 2.707 .333
CONTST (control) 66 12.0152 2.754 .339
t = -.65 p =.517        
Variable Number of Cases   Standard Deviation Standard Error
POPTST (popular) 64 11.8438 2.779 .347
CONTST (control) 64 11.9375 2.822 .353
t = -.33 p =.742        

Table 6 presents the results of the post-hoc procedure where the variables YC ("When I study, listening to classical music interferes with my concentration") and YP ("When I study, listening to popular music interferes with my concentration") were coded 1 for a yes and 0 for a no response. 33% of the respondents indicated that classical music interferes with their studying, while 43% of the subjects indicated that popular music interferes with their studying. The variables CLASTST and POPTST were then regressed on YC and YP respectively, Controlling for CONTST . The variables YC and YP also had no statistically significant effect on, or correlation with, CLASTST and POPTST respectively. Also, the various independent variables were tested for interactions and curvilinearity, revealing no statistically significant results as illustrated in Table 7.

Table 6

Post Hoc Results of the Effects of Students' Preferences for Study with Music on Math Test Scores

Dependent Variable = POPTST
  DF Sum of Squares Mean Square
Regression 2 136.36675 68.18338
Residual 46 163.87814 3.56257
R Square = .45419   F = 19.13883 Sig F <.0001
Variable B SE B Beta t Sig. t
YP -.451287 .545324 .090680 -.828 .4122
CONTST .665412 .110802 .658044 6.005 .0000
(Constant) 4.300081 1.469611      
Dependent Variable = CLASTST
  DF Sum of Squares Mean S!guare
Regression 2 135.04316 67.52158
Residual 46 165.20174 3.59134
R Square = .44978   F =18.80121 Sig F <.0001
Variable B SE B Beta t Sio@'. t
YC .330808 .593371 .061592 .558 .5799
CONTST .666549 .111715 .659168 5.967 .0000
(Constant) 3.981855 1.420960      

Table 7

Correlation Coefficients of Classical and Popular Music Listening Styles

CLASTST 1.0              
CA -.09 1.0            
CC -.08 .59 1.0          
CP -.09 .73 .50* 1.0        
CT -.14 .35 .19 .20 1.0      
CE -.11 .77 .59* .78* .27* 1.0    
YC .16 .08 -.01 -.15 -.02 -.21 1.0  
CONTST .71 .20 -.18 -.10 -.10 -.13 .07 1.0
CLASTST CONTST CA CC CP CI CE YC  
POPST 1.0              
PA -.12 1.0            
PC -.04 .36 1.0          
PP -.01 .66 .46* 1.0        
PI -.09 -.05 -.03 .01 1.0      
PE -.06 .77 .45 .75* -.07 1.0    
YP -.18 -.02 .27 -.02 .16 -.17 1.0  
CONTST .67 -.13 -.04 .01 -.11 -.10 -.11 1.0
POPT ST CONTST PA PC PP PI PE YP  

p <.05

 

Conclusions and Discussion

The music listening styles selected as independent variables for this study had no statistically significant effect on the dependent variables, math test scores. Examination of the correlation matrices in Table 7 revealed no significant bivariate correlations between music listening styles and math test scores. However, correlations among the independent variables implied that multicolinearity may have contributed to the lack of statistical support for the model. However, an exploratory factor analysis of the variables did not reveal any second order factors.

CONTST, the math test taken with no background music, was a statistically significant predictor of math performance under both classical and popular music listening conditions. However, there was not a statistically significant difference in the mean scores of the three tests, leading one to question if background music has any effect, either positive or negative, on math test performance. Also, the students' self report data on interference by background music suggested no effect on the math test scores. This finding supports studies by Wolfe (1983) and LaBach (1960) reporting that students' self reports of background music interference had no effect on their performance.

Future studies of this nature might employ a hypothesis suggested by Roth (1975) when he investigated the effects of background music on a variety of academic tasks; and look toward an interaction between the music and task. The present study suggests that music listening styles have no effect on the interference effect of background music. Also students' attitudes toward background music during testing situations may present an interesting.study. Although several studies have refuted the short term effect of background music on academic performance (LaBach, 1960; Wolfe, 1983; Kelly, 1993), a long term effect in attitude toward a subject and achievement might be the subject of future studies.

References

Cox, M. 0. (1981). Effects of hypnotherapy and relaxation training on mathematics achievement. (Doctoral Dissertation, Texas A&M University, 1981). Dissertation Abstracts IntemationaL 42/10, 4186.

Etaugh, C., & Michaels, D. (1975). Effects on reading comprehension of preferred music and frequency of studying to music. Perceptual and Motor Skills, 41, 553-554.

Friedman, B. (1959). An evaluation of the achievement in reading and arithmetic of pupils in elementary school instrumental music classes. (Doctoral Dissertation, New York University, 1959). Dissertation AbstractsInternational, 60, 3662-A-3663A.

Greenberg, R.P., & Fisher, S. (1971). Some differential effects of music on projective and structured psychological tests. Psychological Reports, 28, 817-818.

Hanshumaker, J. (1980). The effects of arts education on intellectual and social development: A review of selected research. Bulletin of the Council for Research in Music Education, 61, 10-28.

Hedden, S. K. (1971). A multivariate investigation of reaction profiles on music listeners and their relationships with various autochthonous and experiential characteristics. (Doctoral Dissertation, The University of Kansas, 1971). Dissertation Abstracts International, 32 2120A.

Henderson, M. T., Crews, A., & Barlow, J. (1945). A study of the effect of music distraction on reading efficiency. Journal of Applied Psychology, 29, 313-317.

Kelly, S. (1993). A comparison of the effects of background music on the reading comprehension of university undergraduate music majors and nonmusic majors. Southeastern Journal of Music Education, 5, 86-97.

LaBach, J. P. (1960). The effects of background music on reading comprehension and various other characteristics in sixth grade students. (Doctoral Dissertation, Syracuse University, 1960). Dissertation Abstracts International, 21, 350-A-351-A.

Madsen, C. K. (1987). Background music: Competition for focus of attention. In C. K. Madsen & C. A. Prickett (Eds.), Applications of research in music behavior (pp. 315-325). Tuscaloosa, AL: The University of Alabama Press.

Madsen, C. K., & Forsythe, J. L. (1973). Effect of contingent music listening on increases of mathematical responses. Journal of Research in Music Education, 21, 176-181.

Moller, L.E. (1980). Performance of musicians under noise. Perceptual and Motor Skills, 50, 301-302.

Radocy, R. E., & Boyle, J. D. (1988). Psychological foundations of musical behavior (2nd ed.). Springfield, IL: Charles C. Thomas.

Roth, C. F., Jr. (1975). The effects of background music on knowledge, comprehension, application, and analysis types of learning. (Doctoral Dissertation, Southern Illinois University, 1975). Dissertation Abstracts International, 36, 7751-A.

Stewart, R. E. (1984). A comparison of the music preferences of students in three educational systems: Seventh-day Adventist, public, and private independent schools. (Doctoral Dissertation, The University of Miami, 1984). Dissertation Abstracts International, 45, 1066A

Traver-Holder, A. (1993). At-risk students and math achievement: The effects of manipulative-based and music-based teaching approaches. (Masters Thesis, Regent University, 1993). Masters Abstracts International, 31/04, 1440.

Wolf, R. H., & Weiner, F. F. (1972). Effects of four noise conditions on arithmetic performance. Perceptual and Motor Skills, 35, 928-930.

Wolfe, D. E. (1983). Effects of music loudness on task performance and self-report of college-aged students. Journal of Research in Music Education, 31,191-201.

Wolff, K. 1. (1977). The nonmusical outcomes of music education: A review of the literature. Bulletin of the Council for Research in Music Education, 55,1-27.