by TIM METCALFE
Background
A growing body of research supports an influence of background music upon myriad motor-behavioural outcome measures. This effect is described as tendentially positive, in that motor behaviour is typically augmented in some sense by background music (Kämpfe, Sedlmeier, & Renkewitz, 2011). For example, Milliman (1982) found that increasing the tempo of background music (130 bpm compared to 70 bpm) in a supermarket environment led to a significant increase in the average pace of shopper traffic. Guéguen, Hélène and Jacob (2004) demonstrated that musical loudness may also affect the speed of motor behaviour: the researchers observed that louder background music (88-91dB SPL compared to 72-75dB SPL) led to patrons in a bar consuming drinks at a significantly faster rate. Using a treadmill-running paradigm, Edworthy and Waring (2006) demonstrated that musical tempo may exert a greater influence upon running speed when the music is louder. Thus, empirical research has shown that changes in specific musical variables, and interactions thereof, may be associated with specific motor-behavioural outcomes. More recently, it has been suggested that background music might also affect the vigour with which motor behaviour is performed; in a synchronised walking paradigm, Leman et al. (2013) found that, with background musical tempo held constant, louder music tended to elicit greater average stride lengths.
Effects of background music upon motor behaviour constitute an important corpus of findings because background music is increasingly ubiquitous and relatively easy to manipulate (Milliman, 1986). However, in order for the manipulation of musical variables to be maximally impactful, research must strive towards a thorough understanding of the cognitive mechanisms underlying the musical-motor relationship. To this end, two major theories have been postulated: the arousal hypothesis (Husain, Thompson, & Schellenberg, 2002) and the entrainment hypothesis (Brodsky, 2002). The former asserts that music increases physiological arousal which leads to greater motor-behavioural speed, while the latter suggests that the effect arises due to temporal synchronisation of behaviour with musical stimuli. Unfortunately, neither theory has, as yet, been able to fully account for the effects of music on motor behaviour.
This study aimed to build upon the existing literature by more comprehensively investigating the effects of different musical variables on motor behaviour. The study investigated the effects of tempo, loudness, and a novel variable: jitter – i.e. weaker temporal regularity caused by increased standard deviation of interonset intervals (Sievers, Polansky, Casey, & Wheatley, 2013). These variables were chosen in order to tease apart the arousal and entrainment accounts – e.g. effects of jitter may be explained more parsimoniously by the latter hypothesis, since jittered stimuli are more difficult to synchronise with. Additionally, prior research has not always been optimally controlled. In manipulating tempo for example, various studies have introduced potentially confounding changes to other variables e.g. genre (Brodsky, 2002). Therefore, it was important for the current study to maintain maximum control over all musical variables in order to rule out such confounds.
Aims
The study aimed to replicate existing effects of musical tempo on motor behaviour and to investigate additional musical variables, loudness and jitter, in order to evaluate the arousal and entrainment hypotheses. Further, the study sought to investigate potential interactions between these variables and to examine the extent to which effects of musical variables upon motor behaviour are mediated by concurrent changes in arousal state. Using this knowledge, the study set out to establish an integrated arousal/ entrainment theory of how music influences motor behaviour, and to suggest in greater detail how background music might be employed practically.
Method
Participants
Thirty University of Sheffield students took part voluntarily (14 male, 16 female; mean age = 25.67, SD = 6.80) after giving informed consent.
Procedure
A simple walking paradigm was used, in which participants walked along a path at a natural pace for 250m per trial, while listening to systematically varied background music. A smartphone, held by participants, was used for playback of the stimuli, with noise-cancelling headphonesused to eliminate inter-trial variability in ambient noise. A pedometer application (Runtastic Pedometer Pro) was used to record participants’ walking speed, step rate and stride length. Subjective arousal state and liking of the musical stimuli were also measured via a brief questionnaire adapted from van der Zwaag, Westerink and Broek (2011). Participants were randomly assigned to either ‘loud’ or ‘quiet’ stimuli conditions, and completed five trials each: four experimental conditions and a baseline (no music) condition. The baseline always came first, whilst the order of the other conditions was randomised.
Stimuli
Stimulus music was composed electronically using digital audio workstation software (FL Studio 10). To preserve ecological validity, the piece was designed to closely resemble contemporary pop/ dance music. The piece was composed in 4/4 time, in D-flat major, and featured: familiar instruments (piano, synthesizers etc.), simple melodies and a conventional verse/ chorus structure. The beat was clearly demarcated throughout by a kick drum occurring on every beat of the bar with equal intensity. The piece was entirely instrumental, to exclude the possibility of confounding lyrical/ extra-musical associations.
Eight different versions of the stimulus were created, featuring manipulations of: tempo (70 bpm vs. 130 bpm), loudness (60dB SPL vs. 80dB SPL) and jitter (jitter vs. no jitter). The jitter manipulation was achieved by exporting instrumental tracks as MIDI, converting to text, and adding a randomly-generated number between -5% to 5% of each inter-onset interval to its original value. This range was chosen to be rhythmically disruptive, without drastically affecting melodic structure.
Results
Prior to analysis, inter-individual differences were attenuated by subtracting each participant’s baseline data for each measure from the corresponding data for each measure in the experimental conditions.
ANOVA revealed a significant main effect of musical tempo upon walking speed in ‘fast’ and ‘slow’ conditions F(1, 28) = 24.24, p < .001, r = .46. Specifically, walking speed was greater with faster music, compared to both slower music and no music (Figure 1). There was also a significant interaction between tempo and jitter, F(1, 28) = 7.23, p = .012, r = .21. This remained significant when controlling for the influence of arousal and musical preference using regression: tempo explained a significant proportion of variance in walking speed in the non-jitter condition, β = .56, t(59) = 3.89, p < .001, but not the jitter condition. This suggests that the interaction may not be attributed merely to corresponding changes in arousal/ mood. Importantly, the music influenced arousal/ mood factors as expected (e.g. fast, non-jittered stimuli were associated with greater positive feelings), ruling out the idea that the lack of a mediating role for arousal denoted a failure of the musical stimuli to elicit changes in arousal state.
The effect of loudness upon walking speed, although in the direction predicted by the arousal hypothesis (i.e. louder music led to faster walking), was not statistically significant.
Figure 1. Participants’ mean walking speed (as difference from baseline), per condition. Error bars denote standard error of the mean.
The average number of steps per minute taken by participants was also affected significantly by musical tempo, F(1, 28) = 4.84, p = .036, r = .15, though the interaction between tempo and jitter missed significance F(1, 28) = 3.71, p = .064, r = .12. The effect of tempo upon average step rate was not significant once arousal and preference covariates were included, however.
Participants’ maximum number of steps taken per minute was also influenced significantly by tempo, F(1, 28) = 8.15, p = .008, r = .23, and tempo explained significant variance in maximum step rate even with arousal and preference covariates included, β = .36, t(119) = 3.47, p = .001. Effects of jitter, loudness and all interactions, however, were non-significant.
Stride length was affected significantly by musical loudness, F(1, 28) = 6.76, p = .015, r = .19, but not by tempo or jitter. Loudness was able to explain a significant proportion of variance in stride length, even when arousal and preference covariates were included, β = .39, t(119) = 4.51, p < .001.
Conclusions
The study was successful in replicating the effects of musical tempo upon motor-behavioural speed, and in extending these effects to include the influence of jitter. Where stimulus pulse clarity was reduced by a jitter manipulation, walking speed was influenced significantly less by musical tempo. Since this relationship remained statistically significant when controlling for concurrent changes in arousal level, it may be concluded that the addition of jitter exerted an effect primarily by disrupting the process of musical-motor entrainment. It should be cautioned however, that this constitutes only indirect evidence for the entrainment hypothesis – since average step rates did not match with stimuli tempo, it cannot be claimed conclusively that a process of synchronisation took place. A more sophisticated analysis of step rate over time may be valuable in further assessing this discrepancy.
Elsewhere, tempo affected step rate as expected, though jitter exerted less influence than on overall walking speed. Musical loudness primarily affected stride length, which supports previous research (Leman et al., 2013). Unexpectedly however, this effect appeared not to be driven by arousal. Potentially, this might be explained in terms of inconsistency between subjective and physiological measurements of arousal.
To best explain the observed pattern of results, an integrative model is proposed, in which the function of arousal is to direct attentional resources towards rhythmic stimuli, facilitating entrainment. Thus, once arousal meets some threshold and the external stimulus is appraised as being potentially ‘synchronisable with’, increasing arousal further should make little difference. This might explain why arousal appeared to exert little modulatory influence in the current study. Though this account remains speculative at present, it is hoped that it may provide an appropriate framework to facilitate future research.
With respect to the practical usage of background music to modify motor behaviour, the study suggests that music should be loud enough to direct people’s attention towards it and should contain as clear a pulse as possible in order to maximise potential for synchronisation. Assuming that the volume is sufficient to facilitate the direction of attention, further increases in loudness appear to affect motor vigour as opposed to behavioural speed. Future research should clarify the extent to which musical tempo must correspond to the existing motor behavioural speed, in order to affect a change – effectively establishing a ‘window of optimal musical-motor synchronisation’, i.e. a range either side of the typical motoric speed (for a given behaviour), by which synchronisation is easily achievable. This type of research, combined with the results of the current study, should maximise the efficacy with which motor behaviours may be wilfully manipulated in practical settings.
Notes
Address for correspondence: Tim Metcalfe, Department of Music, University of Sheffield, 34 Leavygreave Road, Sheffield, S3 7RD.
Email: tmetcalfe1@sheffield.ac.uk.
References
Brodsky, W. (2002). The effects of music tempo on simulated driving performance and vehicular control. Transportation Research, Part F, 4, 219-241.
Edworthy, J., & Waring, H. (2006). The effects of music tempo and loudness level on treadmill exercise. Ergonomics, 49(15), 1597-1610.
Guéguen, N., Hélène, l. G., & Jacob, C. (2004). Sound level of background music and alcohol consumption: an empirical evaluation. Perceptual and Motor Skills, 99(1), 34-38.
Husain, G., Thompson, W. F., & Schellenberg, E. G. (2002). Effects of musical tempo and mode on arousal, mood, and spatial abilities. Music Perception, 20(2), 151-171.
Kämpfe, J., Sedlmeier, P., & Renkewitz, F. (2011). The impact of background music on adult listeners: a meta-analysis. Psychology of Music, 39(4), 424-448.
Leman, M., Moelants, D., Varewyck, M., Styns, F., van Noorden, L., & Martens, J. P. (2013). Activating and relaxing music entrains the speed of beat synchronized walking. PloS One, 8(7), e67932.
Milliman, R. E. (1982). Using background music to affect the behaviour of supermarket shoppers. The Journal of Marketing, 46, 86-91.
Milliman, R. E. (1986). The influence of background music on the behaviour of restaurant patrons. Journal of Consumer Research, 13(2), 286-289.
Sievers, B., Polansky, L., Casey, M., & Wheatley, T. (2013). Music and movement share a dynamic structure that supports universal expressions of emotion. Proceedings of the National Academy of Sciences, 110(1), 70-75.
van der Zwaag, M. D., Westerink, J. H. D. M. & van den Broek E. L. (2011). Emotional and psychophysiological responses to tempo, mode and percussiveness. Musicae Scientiae, 15(2), 250-269.