Musical information increases physical performance for synchronous but not asynchronous running

Abstract

Given that physical performance is enhanced by listening to music, what information in the music is the active ingredient? Here, we varied the amount of music information in an otherwise identical piece of music, from only the rhythm, through a synthesized and scaled down version, to the full original version. Twenty-two university students (11 males and 11 females) ran for 10 minutes to each of eight conditions, two with white noise, three with music that facilitated synchronization with the running pace, and three with tempi where synchronization was impossible. Dependent variables were distance run and the number of steps, from which stride length was computed. Heart rate and mood (PANAS) were also measured for control purposes. Participants tended to run a greater distance when there was more music information, which was mainly an effect of longer strides rather than a faster stride rate. This effect was stronger in the synchronous conditions. The results suggest that the motivational effects of music information during running is mostly related to richer temporal information conveyed by faster metrical levels, when attempting to synchronize with the beat in the music

Keywords

asynchronous music mood motivation music information outdoor running synchronous music

Athletes as well as non-professionals engaging in physical exercise frequently use music while exercising. Apart from perceived psychological benefits such as later experienced onset of fatigue, music has also been shown to improve physical performance (Yamamoto et al., 2003). The largest effects in this regard have been found for self-selected music, that is, songs that the participants have chosen themselves (Biagini et al., 2012). The importance of using motivational music to enhance performance has attracted considerable attention in recent decades (Priest, Karageorghis, & Sharp, 2004), but the studies have, however, not identified what it is about the music that induces these effects. It could be anything from mere familiarity with the music and its extra-musical associations, over features that have idiosyncratic effects on a particular individual or physical exercise (Madison, Paulin, & Aasa, 2013; Sanchez, Moss, Twist, & Karageorghis, 2014), to general psychophysiological effects of rhythmic or harmonic aspects of the musical structure (Madison et al., 2013). It is difficult to consider musical structure because music is a multifaceted stimulus that also changes over time. There are hence innumerable conditions to compare before we can determine their relative effects, which would be very impractical. The approach taken here is known as copy synthesis, meaning that we take an ecological stimulus and create systematically different versions of it (Madison, 2000). Given that previous research has typically compared music with silence and with entirely non-musical stimuli, such as white noise, we will here compare stimuli with more or less musical content; that is, in which the amount of musical information is varied parametrically.

Karageorghis and colleagues identified four factors that predict participants’ psychophysical response to music, namely cultural impact, musicality, rhythm response, and association with extra-musical thoughts, feelings and images that the music may evoke (Karageorghis, Terry, & Lane, 1999). The effects of the tempo aspect of musical rhythm on various forms of sports performance have been explored (Karageorghis et al., 2011; Karageorghis, Jones, & Stuart, 2008), and music with which one can synchronize one’s limb movements has proved particularly effective for enhancing exercise performance (Bacon, Myers, & Karageorghis, 2012; Bood, Nijssen, van der Kamp, & Roerdink, 2013). An important aspect of the music and performance relationship is therefore whether movement is synchronized with the pulse in the music or not. No studies have, however, explored the role of musical components or information in relation to whether participants synchronize or not. Because the performance-enhancing mechanisms may be quite different across synchronous and asynchronous motion, it is important to consider them separately in relation to musical structure.

Mood is another aspect believed to mediate the association between exercise performance and music (Seath & Thow, 1995). Reported increased pleasant and decreased unpleasant emotions were associated with higher performance during running while listening to music (Lane, Davis, & Devonport, 2011). It should also be noted that almost all studies that involve emotions rely on self-ratings, which are prone to several problems of interpretation, including subjectivity in assessing one’s own state of mind based on cognitive and other factors, confounding variables, and common method variance. Nonetheless, it seems important to at least attempt to control for possible effects of emotions and mood when studying effects of music on physical performance. In conclusion, there are several strands of research that suggest a strong link between music and athletic performance, but the current state of research does not indicate why this is the case or which mediating psychological factors are involved.

It has been shown that there is an association between exercise performance and musical rhythm (Karageorghis, Terry, & Lane, 1999) but it is unclear what information in the music is the ‘active ingredient’. It is arguable that whatever effects that derive from the musical structure itself are related to the amount of structural information. For example, if there were no change in physical performance between a typical piece of music and a version of the same piece stripped of melody, harmony, and rhythmic embellishments, this would indicate that the remaining information alone is sufficient for whatever function the music has in this context. A dose-response relationship between performance and the amount of structural information would however indicate that this information is used by the neural system. Such a distinction would have important implications for the nature of the music-physical performance relationship, and help form hypotheses to further investigate the precise mediating mechanisms.

The aim of the present study is to investigate to what extent musical information per se has this motivating or enhancing effect. We experimentally tested the effect of the amount of music information on running performance. To include the possibility that even listening to a regular pulse increases performance, we also included a condition with white noise. Defining performance as running distance, we hypothesized that (1) synchronization yields higher performance than non-synchronization, and (2) more music information leads to increased performance regardless of synchronization. Possible interactions between these two variables remain an open question.

Method

Design

The study comprised a 2 (synchronization) × 4 (music information) within-participants design, with its basic features depicted in Figure 1. Synchronization was manipulated by the tempo of the music, which was chosen to correspond to each participant’s actual running pace in the synchronous condition, and to be at least 30% slower than this tempo in the asynchronous condition. Music information occurred in four levels, three music-like and one with white noise, as described under the Music selections heading below. The eight conditions were delivered in a balanced and rotated order across participants, which eliminated possible order effects. Each session included one music condition in both synchronous and asynchronous versions, which made it necessary to have both white noise conditions in the same session. The three blocks with music-like stimuli began either with synchronous or asynchronous music, the order of which was counter-balanced such that each participant had one session with one order and two sessions with the other order, which eliminated order effect across participants.

Figure 1.

Overview of the design and data collection process.

One advantage of running as a model behaviour is that it is simple and natural, and thus has high ecological validity, but a disadvantage is that it is unconstrained, inasmuch as performance can be altered in different ways. The distance covered is the product of stride length and stride rate, for example, and may also interact with exertion. To control for these factors, we measured both velocity (in terms of time to complete a fixed distance), the number of strides taken, and the heart rate (HR). HR was also used to ascertain that participants ran within 65%–75% of maximum capacity. This range is optimal because it is demanding enough to test behavioural influences to but still not so hard as to produce lactic acid that might decrease performance (Hayakawa, Miki, Takada, & Tanaka, 2000). In contrast to previous running studies (e.g., Bood et al., 2013), we opted for outdoor running because it avoids the confounding cognitive and motivational operations related to adjusting the treadmill speed.

Participants

Participants were recruited through announcements placed on the university webpage and paper flyers handed out manually at a fitness centre close to the university. All participants were students or employees at the university, were a homogenous group in terms of age and fitness levels, and participated in at least 5 km running per week, non-elite running, or in regular sports games. Exclusion criteria were having had illnesses or used any drugs or medications within 3 months before the first session. The sample size was determined by a power analysis based on a pilot study with six participants fulfilling the same criteria who ran to the same synchronous or asynchronous music as well as white noise. With an alpha set at 0.05 and power at 0.80, approximately 21 participants were required to identify a 25 m difference in the distance covered. The pilot study also showed the amount of variation in the running distance on the same track on different days. To account for experimental dropout and noncompliance with the protocol, 35 volunteer participants were recruited. At the end of data collection, 25 participants had completed all running sessions, but only 22 participants were used in the analysis as three of them had not followed instructions. All participants gave written informed consent before commencing the study, which was approved by the Regional Ethical Committee at Umeå University (2012-211-31M).

Music selections

The participants of the pilot study rated the motivational qualities of the 10 popular music compositions used in a previous study (Madison et al., 2013) on the Brunel Music Rating Index (BMRI–II) (Karageorghis, Priest, Terry, Chatzisarantis, & Lane, 2006), using custom-built music rating software that played the music and recorded the participants’ ratings. Two music tracks were chosen on the basis that both were high and identical in BMRI ratings (M = 6.3), that one was close to running tempo, and the other one so different in tempo that it would be almost impossible to synchronize with when running. One track was ‘Stop the rock’ (Apollo 440, 1999, track 2; 155 BPM) and the other was ‘The big jump’ (The Chemical Brothers, track 6; 119 BPM). The former has a tempo close to natural running pace, which could be adapted in tempo to each participant’s individual natural running pace. ‘Stop the rock’ is thus henceforth referred to as the synchronous music and ‘The big jump’ as the asynchronous music. The original versions of these tunes were presented as-is for the highest music information levels, S4 (synchronous) and AS4 (asynchronous). Consistent with this numbering, S2, S3, AS2, and AS3 versions were synthesized using Cubase (Steinberg AG) to produce alternative versions with less information than the original sound track, as described in more detail in (Madison et al., 2013). Music information level 2 consisted of a simplified version of what the drum kit was playing in the original version, and level 3 was a synthesized performance with all the defining features of the composition, including the drum part, bass, harmony, and melody, all rendered with similar and realistic synthesizer sounds. All the three synchronous music conditions were time-stretched in the software Audacity to make their tempo equal to the spontaneous running tempo of each individual participant, as measured in the habituation trials (Bood et al., 2013).

Positive Affect Negative Affect Scale (PANAS)

The participants filled out a Positive Affect Negative Affect Scale (PANAS) questionnaire (Tellegen, Watson, & Clark, 1988) before and after each running trial, including the habituation trials. PANAS measures an individual’s affective response in a given situation, and is considered to be closely associated with an individual’s mood. Positive affect is associated with being energetic and cheerful, and Negative affect is about being distressed and upset (Crawford & Henry, 2004).

Procedure

Prior to the experiment proper, each participant met with the test leader for one habituation session, during which the participant was given careful instructions regarding the equipment and the task, was notified that only two trials were to be run, and filled out the PANAS and an inclusion and exclusion criteria questionnaire. The average number of steps recorded across the two runs was used to select the individual tempo for the synchronous music for each participant. The following four sessions were scheduled in the morning before lunch, and at the same time for each participant, to minimize possible effects of diurnal variation. Participants were instructed to eat the same breakfast at least 2 hours before the experiment on all session days, and were required to refrain from to tobacco products, alcohol, and intense exercise and other intensive physical activity on the day before the session. The participants were not given specific instructions on how fast to run or if they should synchronize to the music, but were asked to do what felt natural.

A 300 metre outdoor track was prepared in an open expanse close to the campus. Each session started with attaching a Polar RCX5 RUN wrist watch with a radio signal connected chest belt and a Polar S3+ foot sensor that was attached to the shoelace of the right shoe. This device measures both heart rate and stride frequency, and stores these data in the wrist watch. The participant then warmed up by running while the equipment and its set-up was tested. Following this the participant ran for two 10-minute trials separated by a 20-minute break. Each trial began by the participant starting the mp3 player, which issued an instruction to start the recording on the watch and to wait and be prepared, followed by a countdown towards when to start running, at which point the music started and continued for exactly 10 minutes. The music was delivered by a Philips Go Gear Raga mp3 Player and high-definition headphones. The experimenter started the stopwatch when the participant started running, and made sure that the participant stopped the HR recording when the 10 minutes had passed.

After that the participant indicated their level of exertion on a Borg scale chart that the experimenter held up. Pre- and post-test questionnaires were also filled out before and after each running trial. The participants were allowed to relax during the 20-minute break and to drink up to 20 centilitres of water. Participants were not allowed to run with a cold, cough, flu, other infections, or if they were on medication, in which case the next session was rescheduled. The weather is known to affect running and hence any scheduled sessions on rainy days were also rescheduled.

The data from the Polar watch was transferred to a computer after each session, and the total number of steps was calculated based on the number of steps and the distance covered.

Statistical analysis

The data were screened for outliers, on which basis three participants were excluded from further analyses because they seemed not to have followed instructions, and one of them had also covered a very short distance. For PANAS, the difference between pre- and post-running was used in the following analyses. The average heart rate was obtained from the heart rate sensor. Data were analysed using a series of repeated-measures analyses of variance with distance, steps per minute and stride length as dependent variables and the music information conditions (of both synchronous and asynchronous) and white noise as the independent variables.

Results

Of the 22 participants that remained after excluding outliers, 11 were males (age M = 25.0; SD = 4.84) and 11 females (M = 24.7; SD = 2.91). The descriptive statistics of the participants’ performance in each condition are presented in Table 1.

Table 1.

Descriptive statistics of participants’ performance for each condition and sex.

Variable	Music condition	Male	Female	All
		n = 11	n = 11	N = 22
Distance (m)	AS4	2239.2 (237.1)	2015.4 (277.1)	2146.9 (236.2)
	AS3	2112.3 (297.7)	2047.6 (169.5)	2119.4 (231.4)
	AS2	2239.2 (255.2)	1995.3 (214.1)	2078.4 (256.0)
	S4	2282.6 (328.7)	2003.1 (291.2)	2191.8 (290.9)
	S3	2108.4 (387.6)	2025.3 (298.6)	2184.5 (302.5)
	S2	2054.9 (232.3)	2059.2 (410.9)	1999.9 (254.7)
	WN	2044.9 (364.8)	1891.5 (252.1)	1951.9 (293.4)
Stride length (cm)	AS4	132.0 (10.5)	123.7 (18.1)	127.8 (15.0)
	AS3	131.7 ( 18.0)	126.4 (10.2)	129.0 (14.5)
	AS2	131.1 (12.3)	118.7 (10.7)	124.9 (12.9)
	S4	134.2 (19.7)	132.5 (15.2)	133.3 (17.2)
	S3	132.2 (16.8)	132.9 (10.8)	132.5 (13.8)
	S2	124.3 (16.5)	115.4 (17.6)	119.8 (17.3)
	WN	124.1 (18.4)	112.3 (10.1)	118.2 (15.9)
Average heart rate (BPM)	AS4	174.9 (15.5)	181.4 (9.7)	178.2 (4.6)
	AS3	173.9 (15.6)	183.7 (11.6)	178.8 (6.9)
	AS2	179.4 (10.1)	181.0 (11.8)	180.2 (1.1)
	S4	174.7 (14.2)	186.8 (17.4)	180.8 (8.6)
	S3	178.1 (15.1)	182.2 (11.8)	180.2 (2.9)
	S2	173.6 (11.4)	180.0 (10.7)	176.8 (4.5)
	WN	160.2 (13.3)	164.9 (9.8)	162.6 (3.3)
Positive affect difference	AS4	1.9 (6.8)	1.7 (5.7)	1.8 (0.1)
	AS3	2.1 (4.5)	1.2 (3.1)	1.65 (0.6)
	AS2	2.0(6.1)	1.4 (3.9)	1.7 (0.4)
	S4	1.9 (4.2)	4.8 (4.2)	3.4 (2.1)
	S3	2.0 (4.5)	4.2 (5.0)	3.1 (1.6)
	S 2	1.4 (4.4)	0.9 (6.1)	1.2 (0.4)
	WN	2.0 (4.0)	2.4 (3.9)	2.2 (0.3)
Negative affect difference	AS4	−7.0 (4.9)	−1.1 (2.8)	−4.1 (4.2)
	AS3	−2.0 (4.2)	−0.9 (3.2)	−1.4 ( 0.8)
	AS2	0.3 (2.0)	−1.0 (2.4)	−0.4 (0.9)
	S4	−0.5 (1.2)	−0.9 (2.3)	−0.7 (0.3)
	S3	−2.8 (4.6)	−0.8 (2.9)	−1.8 ( 1.4)
	S2	0.7 (1.1)	0.7 (3.5)	0.0 ( 0.9)
	WN	−0.8 (4.3)	−0.2 (1.7)	−0.5 (0.4)

Note. Stride frequency (pace) is plotted in Figure 2 and is therefore not copied in the table. WN = white noise.

First, we examined to what extent participants actually synchronized in each trial, thus providing a manipulation check of this independent variable. To this end, the differences between the participant’s mean stride frequency (pace) and the beat tempo of the music were computed for each trial. The pace-tempo difference varied from −12.7 to 15.1 BPM for synchronous running (SD = 5.0), and from 25.0 to 62.6 BPM for asynchronous running (SD = 7.7). Figure 2 shows, on the left ordinate, that the pace did not differ systematically across the conditions, except for a tendency to be lower for white noise. The right ordinate is scaled to the mean differences between pace and music tempo for all conditions, and the plot shows that they corresponded almost exactly to each other in all the synchronous conditions, but that participants ran on average about 46 BPM faster than the tempo in all asynchronous conditions. This corresponds of course to the difference between the ~166 BPM mean pace and the 119 BPM tempo in the asynchronous music.

Figure 2.

Mean stride frequency (pace) and difference between pace and music tempo for each condition.

The average heart rate in each running sessions was determined to be below 75% of maximum capacity, according to the formula HR_max = 208 −(0.7 × age). The most direct and natural measure of performance is the distance that the participants ran in the 10 minute interval. The men ran a significantly greater distance (2,140 ± 317 m) than did the women (2,015 ± 227 m). A two-way (2 synchronization × 4 music information) within-participants ANOVA with distance as dependent variable showed a significant effect of music information, F3,63 = 3.08, p = .033, but not of synchronization, F2,21 = 0.19, p = .66, nor their interaction, F3,63 = 0.21, p = .89. Since there was no interaction and no difference in running distance between the two white noise conditions, the latter were aggregated for the following analyses. Figure 3 qualifies the ANOVA results, and shows that although the parametric interaction was non-significant, the pattern of significant effects nevertheless differed between synchronous and asynchronous running, such that for asynchronization there was no significant difference between any information levels, but for synchronization there was a significant difference between information level 2 and 3, according to .95 confidence intervals. When the tempo of the music facilitated synchronization, participants ran significantly longer in music information conditions S3 and S4, as compared to white noise and information condition S2. In contrast, music with a tempo that did not allow synchronization caused a significant increase in distance compared to white noise, but no significant differences as a function of music information.

Figure 3.

Mean distance run as a function of music information condition and synchronization behaviour (asynchronous or synchronous running).

Further explorations were made to determine how participants managed to run further, given they did not increase their pace. Zero-order correlations suggested that stride length was a more important determinant of distance (r = 0.90, p < .00001, N = 176) than was pace (r = 0.26, p < .0005). A multiple regression confirmed this pattern, showing that stride length and pace together explained 94.6% of the variance (88% and 13%, respectively, both p < .00001). Measurement error accounts for the remaining 5.4%. Whether the asynchronous conditions provided more leeway for changing the pace was assessed by separate regressions for synchronization behaviour (presented in Figure 4), showing a slightly higher contribution from pace for the asynchronous conditions (beta pace = 0.427, stride = 0.931) as compared to the synchronous (beta pace = 0.346, stride = 0.930).

Figure 4.

Mean step length (stride) as a function of music information and synchronization behaviour (asynchronous or synchronous running).

Discussion

The present study tested the hypotheses that more music information leads to increased running performance, and that synchronization to the beat in the music also leads to increased performance compared to asynchronous running. The first hypothesis was supported and the second was falsified by the data. An additional finding was, however, that music information had a substantially larger and statistically significant effect for synchronous running than for asynchronous running, although the parametric interaction was not in itself significant. We found that stride length mainly accounted for these effects, while the step frequency (pace) was almost constant across all conditions. There was no difference in this pattern across the two synchronization behaviors, which may seem surprising given that the asynchronous music provided no limitations as to which pace to keep. The explanation is apparently that an individual pace was maintained, quite unaffected by the stimuli. This indicates a strong individual preference for a physiologically optimal locomotion pattern (MacDougall & Moore, 2005; Todd, Cousins, & Lee, 2007). Consistent with this, the range of differences between pace and tempo demonstrate that no precise beat-for-beat synchronization could have occurred throughout the trials. On the other hand, it is possible that precise synchronization occurred for longer or shorter parts of the trials. Since we did not record the time of each step, but rather the total number of steps across a 10-minute interval, this could not be determined. It should be noted for future studies that even though the stimulus provides the precise individual movement tempo, perfect synchronization does not occur during running. This is quite interesting, given that runners often claim that they do synchronize and that they pay a lot of attention to choosing music with an optimal tempo. Either synchronization is in fact not important at all, or the nature of synchronization in this context is different from the perfect beat-to-beat correspondence in music performance. One can speculate that this requirement is relaxed for running and perhaps for whole-body motion in general, such that the phase may wander and even jump a beat now and then, or that it is sufficiently rewarding to be in synch only for periods of the movement sequence.

Regarding mood, we found no influence of either the positive or negative scale of the PANAS, consistent with previous research (Simpson & Karageorghis, 2006). This indicates that physical performance was not mediated by changes in mood elicited by the music. One should consider that the participants had not chosen the tunes themselves, and there may therefore be a large variation in how much they liked them. Vast differences in how participants liked the songs may obscure smaller effects of music information or synchronization. Earlier studies have found that self-selected music is associated with mood changes during performance (Lane et al., 2011) but music self-selection was not a viable option in this study because the synthesis had to be made in a consistent fashion across all conditions and participants.

This pattern of results suggests that music information is more important during synchronization. Why might that be? Previous research has indicated a strong motivational force related to entrainment, in particular to a regular beat (Janata, Tomic, & Haberman, 2012; Madison, 2006; Merker, Madison, & Eckerdal, 2009). The satisfaction of entraining will obviously not occur for the asynchronous conditions. That there was no substantial difference in performance between synchronous and asynchronous conditions can mean that the asynchronous music was motivating in a different way, perhaps by increasing arousal (Copeland & Franks, 1991; Thompson, Schellenberg, & Husain, 2001).

The results of this study are novel in terms of assessing running performance as a function of music information. Also, few studies on music and performance have employed outdoor running. However, some features of the present study should be considered in interpreting the results. Participants constituted a convenience sample, which has the general limitation that there might be some condition-participant interactions that limit the generalizability of the results. Although all participants fitted into the inclusion criteria on health and training status, there were differences in fitness-related and other more general measures. They were also not informed in advance that the music would be different on different days, so the sound they heard might initially have distracted them. However, this reaction may wear off after the first experimental session, after which they would expect to run to two different music conditions in future sessions. As stated in the opening section of this article, we assumed that music information may exert a general effect on physical performance, although we do not speculate about the precise nature of such effects. This seems unlikely given that music information had no effect in the asynchronous conditions. However, it has been suggested that the ability of the auditory signal to guide accurate synchronization performance is substantially motivating for movement (Davies, Madison, Silva, & Gouyon, 2013; Madison, Gouyon, Ullén, & Hörnström, 2011). The functional explanation is that temporal information, in this case consisting of music events that reflect faster metrical levels in the rhythmic pattern, facilitates accurate and precise synchronization (Madison, 2014). This is consistent with the increase in performance when the full musical structure was included in music information condition 3, but that no further increase was found for the original version level 4, which did not add any structural information but included lyrics and all sorts of human variation that would reasonably constitute emotionally motivating information.

In conclusion, the motivational effects of music information during running are mostly related to richer temporal information conveyed by faster metrical levels, when attempting to synchronize with the beat in the music.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by two grants from the Swedish National Centre of Research in Sports (Centrum för Idrottsforskning) awarded to Guy Madison.

References

Apollo 440. (1999). Stop the rock. On Getting high on your own supply [CD]. Epic.

Bacon

Myers

Karageorghis

(2012). Effect of music-movement synchrony on exercise oxygen consumption. The Journal of Sports Medicine and Physical Fitness, 52(4), 359–365.

Biagini

M. S.

Brown

L. E.

Coburn

J. W.

Judelson

D. A.

Statler

T. A.

Bottaro

. . .Longo

N. A.

(2012). Effects of self-selected music on strength, explosiveness, and mood. The Journal of Strength & Conditioning Research, 26(7), 1934–1938.

Bood

R. J.

Nijssen

van der Kamp

Roerdink

(2013). The power of auditory-motor synchronization in sports: Enhancing running performance by coupling cadence with the right beats. PloS one, 8(8), e70758.

Copeland

B. L.

Franks

B. D.

(1991). Effects of types and intensities of background music on treadmill endurance. The Journal of Sports Medicine and Physical Fitness, 31(1), 100–103.

Crawford

J. R.

Henry

J. D.

(2004). The Positive and Negative Affect Schedule (PANAS): Construct validity, measurement properties and normative data in a large non-clinical sample. British Journal of Clinical Psychology, 43(3), 245–265.

Davies

Madison

Silva

Gouyon

(2013). The effect of microtiming deviations on the perception of groove in short rhythms. Music Perception: An Interdisciplinary Journal, 30(5), 497–510.

Hayakawa

Miki

Takada

Tanaka

(2000). Effects of music on mood during bench stepping exercise. Perceptual and Motor Skills, 90(1), 307–311.

Janata

Tomic

S. T.

Haberman

J. M.

(2012). Sensorimotor coupling in music and the psychology of the groove. Journal of Experimental Psychology: General, 141(1), 54–75.

10.

Karageorghis Jones

Priest

D.-L.

Akers

R.I.

Clarke

Perry

J. M.

. . . Lim

H. B. T.

(2011). Revisiting the Relationship Between Exercise Heart Rate and Music Tempo Preference. Research Quarterly for Exercise and Sport, 82(2), 274–284.

11.

Karageorghis Jones

Stuart

D. P.

(2008). Psychological effects of music tempi during exercise. International Journal of Sports Medicine, 29(7), 613–619.

12.

Karageorghis Priest

D.L.

Terry

P.C.

Chatzisarantis

N.L.

Lane

A. M.

(2006). Redesign and initial validation of an instrument to assess the motivational qualities of music in exercise: The Brunel Music Rating Inventory-2. Journal of Sports Sciences, 24(8), 899–909.

13.

Karageorghis Terry

P.C.

Lane

A. M.

(1999). Development and initial validation of an instrument to assess the motivational qualities of music in exercise and sport: The Brunel Music Rating Inventory. Journal of Sports Sciences, 17(9), 713–724.

14.

Lane

A. M.

Davis

P. A.

Devonport

T. J.

(2011). Effects of music interventions on emotional states and running performance. Journal of Sports Science and Medicine, 10(2), 400–407.

15.

MacDougall

H. G.

Moore

S. T.

(2005). Marching to the beat of the same drummer: The spontaneous tempo of human locomotion. Journal of Applied Physiology, 99(3), 1164–1173.

16.

Madison

(2000). Properties of expressive variability patterns in music performances. Journal of New Music Research, 29(4), 335–356.

17.

Madison

(2006). Experiencing groove induced by music: Consistency and phenomenology. Music Perception 24(2), 201–208.

18.

Madison

(2014). Sensori-motor synchronisation variability decreases as the number of metrical levels in the stimulus signal increases. Acta Psychologica, 147, 10–16.

19.

Madison

Gouyon

Ullén

Hörnström

(2011). Modeling the tendency for music to induce movement in humans: First correlations with low-level audio descriptors across music genres. Journal of Experimental Psychology: Human Perception and Performance, 37(5), 1578–1594.

20.

Madison

Paulin

Aasa

(2013). Physical and psychological effects from supervised aerobic music exercise. American Journal of Health Behavior, 37(6), 780–793.

21.

Merker

B. H.

Madison

G. S.

Eckerdal

(2009). On the role and origin of isochrony in human rhythmic entrainment. Cortex, 45(1), 4–17.

22.

Priest

Karageorghis

Sharp

(2004). The characteristics and effects of motivational music in exercise settings: The possible influence of gender, age, frequency of attendance, and time of attendance. The Journal of Sports Medicine and Physical Fitness, 44(1), 77–86.

23.

Sanchez

Moss

S. L.

Twist

Karageorghis

C. I.

(2014). On the role of lyrics in the music–exercise performance relationship. Psychology of Sport and Exercise, 15(1), 132–138.

24.

Seath

Thow

(1995). The effect of music on the perception of effort and mood during aerobic type exercise. Physiotherapy, 81(10), 592–596.

25.

Simpson

S. D.

Karageorghis

C. I.

(2006). The effects of synchronous music on 400-m sprint performance. Journal of Sports Sciences, 24(10), 1095–1102.

26.

Tellegen

Watson

Clark

(1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063–1070.

27.

The Chemical Brothers. (2005). The big jump. On Push the button [CD]. Astralwerks.

28.

Thompson

W. F.

Schellenberg

E. G.

Husain

(2001). Arousal, mood, and the Mozart effect. Psychological Science, 12(3), 248–251.

29.

Todd

N. P. M.

Cousins

Lee

C. S.

(2007). The contribution of anthropometric factors to individual differences in the perception of rhythm. Empirical Musicology Review, 2(1), 1–13.

30.

Yamamoto

Ohkuwa

Itoh

Kitoh

Terasawa

Tsuda

. . .Sato

(2003). Effects of pre-exercise listening to slow and fast rhythm music on supramaximal cycle performance and selected metabolic variables. Archives of Physiology and Biochemistry, 111(3), 211–214.