Perceptual and Motor Skills, 1995, 81, 435-440.

© Perceptual and Motor SkiUs 1995

RELATIONSHIP BETWEEN HEART RATE AND

PREFERENCE FOR TEMPO OF MUSIC 1

MAKOTOIWANAGA
Hiroshima University

Summary.-Peopleprefer music with tempi ranging from 70 to 100 cycles per

minute similar to that of adults' heart rate within normal daily situations. Previous
studies of the relation between preferred tempi and heart rates using a pure tone also
have indicated that subjects tended to prefer tempi similar to their heart rates. The
present study examined this relationship using a musical piece (the theme of "It's a
small world") as a stimulus. 14 undergraduate women were subjects who searched for
their favorite tempi by controlling the musical tempo by themselves. The most pre-
ferred tempo was close to their cycle of heart beats, however, tempi that were one
and a half and twice as fast as the heart rate were less preferred in the present study
than in a previous study using a tone. Subjects preferred faster tempi in the descend-
ing series of stimuli than in the ascending one, and hence were influenced by the
initial value of the tempo in the trial sequence. The effects due to the differences of
the meaning of the stimuli are considered.

LeBlanc (1982) noted that musical preference was influenced by many

factors. He and his colleagues (LeBlanc, 1981; LeBlanc, Colman, McCrary,
Sherrill, & Malin, 1988; LeBlanc & Cote, 1983; LeBlanc & McCrary, 1983;
LeBlanc & Sherrill, 1986) examined the effects of instrument, vocal vibrato,
musical type, tempo, and aging on musical preference. Of these factors,
musical tempo has been considered the most important factor providing
preference and emotional reactions (Gundlach, 1935; Hevner, 1937) to mu-
sical pieces. Researching the dominant factors affecting preference for tem-
po, therefore, is a readily available strategy for the examination of musical
preferences and responses to pieces.
Fraisse (1982) considered walking pace, pulse rate, and electrical oscilla-
tion of the central nervous system as indexes of natural tempo. Such a na-
tural, spontaneous tempo was called the personal tempo (Temperley, 1963),
and its interval was about 0.75 sec., which was the same timing for adults'
heart beats (pulse) (Marriott, 1977). The hypothesis that heart beat was a
pacemaker of personal tempo was supported by results from Fraisse (1982)
and Wilson (1986) but confounded by those of Dowling and Harwood
(1986) and Temperley (1963). There is no clear conclusion about this hy-
pothesis.
Dowling and Harwood (1986) and Rosenfeld (1985) reported that peo-

'Address enquiries to Makoto Iwanaga, Ph.D., Department of Behavioral Sciences, Faculty of

Integrated Arts and Sciences, Hiroshima University, Kagamiyama 1-7-1, Higashi-hiroshima 739,
Japan.
436 M.IWANAGA

pie tended to prefer musical pieces with tempi ranging from 70 to 100
cycles per minute (cprn). Although they denied there was a relationship be-
tween these tempi and heart rates, these tempi were consistent with the
range of heart beats per minute within adults' daily life activity. -In a series
of studies LeBlanc and his colleagues (LeBlanc, 1982; LeBlanc, et al., 1988).
who searched for preference tempo using a variety of musical pieces as stim-
uli, reported that the younger subjects were, the faster the preferred tempo.
Walter (1983, cited by Buchanan, 1988) reported that children preferred the
same or faster tempi as their heart rates. These findings lead to the assump-
tion that the tempo preference was influenced by the heart rate.
Iwanaga (1995) made subjects search for their favorite tempi by con-
trolling the tempo of a 440-Hz pure tone by themselves. Comparison of
their heart rates with the tempi they chose as preferred tempi indicated that
most preferred tempo was similar to the heart rate, with the next preferred
being one and a half and twice as fast as the heart rate. To extend these
relationships to musical preference, it is necessary to examine the relation-
ship of tempo preference to heart rate using a musical stimulus. The present
study did this by using a musical piece and the same procedure as Iwanaga
(1995).
METHOD

Subjects
Fourteen female undergraduates served as subjects. They ranged in age
from 19 to 22 years. All were volunteers recruited from a class in educa-
tional psychology.
Stimulus
The musical piece chosen as a stimulus was an eight-bars length of the
main theme of Walt Disney's "It's a small world." The stimulus was, repeat-
edly without pause, presented by a synthesizer controlled by a M.I.D.I. (Mu-
sical Instrument Digital Interface) using a personal computer, and its tone
colors were provided by a glockenspiel in the theme and a vibraphone in the
accompaniment within the tone pallet of the computer music system. Pre-
sented tempi could be varied from 30 to 240 cpm. Although the original
tempo of "It's a small world" is 63 cpm, this piece has been commonly
played at various tempi, for example, in arrangements of lullabies, marches,
and so on. Therefore. it was considered adequate as a stimulus.
Procedure
The subject's task was to search for her favorite tempi by changing the
tempo by her own manipulation of the cursor keys of a personal computer.
Subjects were instructed to inform the experimenter that they found their
favorite tempo throughout the sessions. They were allowed to find their fa-
 HEART RATE AND MUSIC TEMPO 437

vorite tempi many times in one stimulus sequence. Sequences of tempo

changes had two directions; that is, tempo varied from slow to fast in an as-
cending series and varied inversely in a descending one (fast to slow). An
ascending and a descending series had three trials, and the order of the se-
ries of trials was counterbalanced between subjects. Because the range of
tempo change was extremely wide, subjects could stop a trial session even
part way through when they judged there could be no preferred tempo in a
stimulus sequence. Durations of stimulus sequences were 3 to 10 min., de-
pending on the subjects' own paces to carry out the task.
Apparatus
Presentation and control of the musical tempo were accomplished with
a Roland Desk-top Music System (50K), a tone module (Roland MT-32),
and a personal computer (EPSON PC-286VE). The stimulus was presented
through a SONY stereoheadphone (DR-4M) and an integrated amplifier
(Alplne/Luxman LV-103u). Heart rate during a task was measured by an
NEC San-ei polygraph 360 system and recorded on a personal computer
(EPSON PC-286LS) by means of an A/D converter.
Data Processing
To examine the relation between heart rate and musical tempo, the pre-
sented tempo of a musical piece was compared with mean heart rate during
the 5 sec. immediately before a subject found her favorite tempo. Because
there were wide individual differences in heart rates, ratios of musical tempi
to heart rates (tempo/heart rate) were calculated to cancel individual differ-
ences. Values of tempi used for the statistical analysis ranged from 50 to 160
cpm (a normal range of musical tempo). As in Iwanaga (995), the propor-
tion of tempo ratios was averaged by steps of 0.17 ratio among subjects. A t
test of the paired-differences was used to examine differences between ratios
of tempi. All data were analyzed by SPSS for Windows, Version 6.0.
RESULTS
Mean heart rates during the search for favorite tempi ranged from 62.6
to 85.2 beats per minute (bprn), The mean for all subjects was 70.8 bpm
(SD =7 .4). Because there were large individual differences in heart rate, pre-
ferred tempi of one subject could not be directly compared with those of
another subject. Preferred tempi were translated to ratios of musical tempi
to heart rates. Therefore, these ratios could be equal despite heart rates be-
ing different among subjects.
Average tempo selected in both ascending and descending series rang-
ing from 50 to 160 cpm was 93.2 cpm (SD=28.3), faster than the original
tempo of "It's a small world." Since mode was 64.0, the most preferred
tempo was little slower than the heart rate. Because the mode was smaller
438 M.IWANAGA

than the mean and skewness of the distribution was .49, the distribution of
preferred tempi was widely expanded to faster tempi. The mean number of
preferred tempi per trial was 6.3 (SD=3.4) for both ascending and descend-
ing series, 6.1 (SD=3.8) for an ascending one and 6.5 (SD=3.3) for a de-
scending one. Table 1 shows the distributions of percentages of tempo ratios
in steps of 0.17 for the combined and for each series separately. A ratio of
1.0 means that preferred tempo is equal to heart rate, and a ratio of 2.0
means preferred tempo is twice as fast as heart rate. In the over-all dis-
tribution presented on the left side of Table 1, the peak was the ratio 1.0,
with a percentage of 18.9. The t tests for paired-differences indicated that
the ratio 1.0 was significantly greater than all other ratios except ratio 1.17
(Ills=3.32 to 12.18,ps<.01 to .000.

TABLE 1
MIlAN PERCENTAGES Of SELECI'ED TEMPI FOR AsCENDING AND
DESCENDING SERIES OF STIMULI AND COMBINED SERIES

Ratio Slimulus Order

Combined Ascending Descending
M SO M SO M SO
0.67 4.46b 5.67 6.89<1 9.80 3.15f 6.34
0.83 11.15b 7.20 1657<1 9.85 6.80£ 8.02
1.00 18.95· 3.99 26.81c 8.55 13.84" 5.21
1.17 15.65 7.02 17.32<1 11.45 15.35" 10,42
1.33 1O.90b 5.07 9.03<1 7.67 12.41" 7.92
150 12.00b 4.98 7.22<1 8.86 14.85" 7.07
1.67 11.09b 5.55 6.05<1 6.23 14.55" 7.37
1.83 7.33b 5.46 5.46<1 6.47 7.65f 7.82
2.00 4.25b 4.45 1.86<1 4.08 6.14 f ·5.55
2.17 2.23b 3.42 0.91d 2.39 3.29f 5.32
2.33 1.45b 3.17 1.57<1 3.23 1.19f 3.41
250 0.54b 1.43 0.31<1 1.12 O.79f 1.98
Note>-« > b. c > d, e > f. resulted from the I tests for paired-differences.

Apparently, however, distributions of tempo ratios are different for the

ascending and the descending series. In the ascending series, the peak of the
distribution was the ratio 1.0 (including 26.8% of the preferred tempi), Com-
pared to other tempo ratios, the ratio 1.0 was significantly greater than all
other ratios (tus=2.27 to 10.22, ps<.05 to .000. On the other hand, the
peak of the distribution in the descending series was unclear. The preferred
tempi ranged from ratios of 1.0 to 1.67; all were almost the same levels of
about 15%. The t tests indicated that tempo ratios from 1.0 to 1.67 were
greater than the other tempo ratios (tIJS = 1.47 to 7.45, ps =.16 to .001).
Tempo ratios of 0.83 (t1J=2.65, p<.05) and 1.00 (/\]=4.33, p<.Ol)
were selected more often in the ascending series than in the descending one.
 HEART RATE AND MUSIC TEMPO 439

On the contrary, percentages of tempo ratios of 1.50 (t 13 =2.38, p<.05), 1.67

(til =3.10, p<.OI), and 2.00 (t 13 =2.21, p<.05) were more often selected in
the descending series than in the ascending one. These results indicated that
subjects tended to prefer the same tempo as their heart rate in the ascending
series, while tempi were shifted to faster tempi than their heart rate in the
descending one.
DISCUSSION
Using a musical piece as a stimulus, subjects tended to prefer tempi
similar to those of their heart rates, despite the limitation of testing only a
sample of young women. These preferred tempi were also influenced by di-
rection of tempo change of the presented musical piece, i.e., by whether the
tempi were presented in ascending (slow to fast) or descending (fast to slow)
order.
The results supported earlier findings of Iwanaga (1995); however,
there were lower percentages at ratios 1.5 and 2.0 than found in the previ-
ous study. This is likely due to the difference in the meaning of the stim-
ulus, since subjects stopped sequences of stimuli when they judged there
would be no further preferred tempi because all remaining tempi were ex-
tremely fast or slow. The meaning of the stimulus imposed a preference for
a limited tempo range. The stimulus used in this study was a musical piece
and precisely metered; however, because the pure tones presented in the'
previous study had less musical meaning, subjects would recognize two or
three successive tones as one group (Zuckerkandl, 1959). Therefore, subjects
were apt to prefer tempi one and a half and twice as fast as their heart rates.
Although the results supported the finding that individuals were apt to
prefer a certain range of tempi such as 70 to 100 cpm, as Dowling and Har-
wood (1986) noted, they did not support LeBlanc's finding that people
prefer faster music (Leblanc, 1982; LeBlanc, et al., 1988). This inconsistency
might be related to the differences in experimental procedures, wherein sub-
jects were forced to judge their preference using the varied tempi of the
same musical piece or of different musical pieces.
Secondly, the distribution of preferred tempi was influenced by the di-
rection of tempo change. Especially in the descending series which started
from an extremely fast tempo, the preferred tempi were widely distributed,
and an apparent peak in preference was not found. Madsen, Duke, and
Geringer (1986) who asked subjects to listen to music with varied tempi re-
ported that subjects preferred the music played at a faster tempo than that
at a slow tempo. In the descending series in this study, because a trial
started with extremely fast tempi (such as 240 cpm), subjects might habitu-
ate to the faster tempi. That is, the preference for tempo is considered to be
influenced by the value of the tempo of the initial stimulus.
This result suggests that the preference for tempo is not only influenced
440 M.IWANAGA

by autonomic activity but also by various other factors as noted by LeBlanc

(1982). It is important to examine the other factors discussed by LeBlanc
(1982) which can affect preference for tempo.
REFERENCES
BUCHANAN, j. c. (1988)An exploratory study of preschool children's synchronization of a se-
lected rhythmic activity"with music set at their heart rates. Unpublished Ph.D. disserra-
tion, Univer. of South Florida.
DOWLING, W. J.. & HARWOOD, D. L. (1986) Music cognition. New York: Academic Press.
FRAISSE, p. (1982) Rhythm and tempo. In D. Deutsch (Ed.), The psychology 0/ music, New
York: Academic Press. Pp. 148-180.
GUNDl.ACH, R. H. (935) Factors determining the characteristics of musical phrases. American
Journal 0/ Psychology, 47, 624-643.
HEVNER, K. (1937) The affective value of pitch and tempo in music. American Journal 0/
Psychology, 49, 621-630.
[WANAGA, M. (1995) Harmonic relationship between preferred tempi and heart rate. Perceptual
and MotorSkills, 81, 67-71.
LEBLANC, A. (1981) Effects of style, tempo, and performing medium on children's music pref-
erence. Journal 0/ Research in Music Education, 29, 255-270.
LEBLANC, A. (1982) An interactive theory of music preference. Journal 0/ Music Therapy, 19,
28-45.
LEBLANC, A.. COLMAN, J., MCCRARY, J" SHERRIl.L, C., &MALIN. S. (1988) Tempo preferences of
different ages of music listeners. Journal 0/ Research in Music Education, 36, 156-168.
LEBLANC, A., &COTE, R. (983) Effects of telTlPo and performing medium on children's music
preference. Journal 0/ Research in MusicEducation, 31, 57·66.
LEBLANC, A., & MCCRARY, J. (1983) Effect of tempo on children's music preference. Journal 0/
Research in Music Education,3l, 283·294,
LEBLANC, A., & SHERRILL, C. (1986) Effect of vocal vibrato and performer's sex on children's
music preference. Journal 0/ Research i11 MUJic Education, 34, 222-237.
MADSEN, C. K., DUKE, R. A" & GERINGER, J. M. (1986) The effect of speed alterations on
tempo note selection. Journal 0/ Research in MUJic Education, 34, 101-110.
MARRIOTI, H. L j. (1977) Practical electrocardiography. Baltimore, MD: Williams & Wilkins.
ROSENfELD, A. H. (1985) Music, the beautiful disturber. Psychology Today, 19(12),48·56.
TEMPERLEY, N. M. (1963) Personal tempo and subjective accentuation, Journal 0/ Ge11eral
PJychology, 78, 267-287.
WILSON, F. R. (1986) Tone deafand all thumbs? New York: Viking Press.
ZUCKERKANDL, V, (]959) The sense 0/ music. Princeton, NJ: Princeton Univer, Press,

Accepted August 8, 1995.