The influence of music on affective, social, and motivational experiences during circuit training in physical education

The role of music in physical education (PE) is underexplored. This study investigated the effects of self-selected versus externally provided music on affect, emotions, social connectedness, and intrinsic motivation during PE lessons. A clustered cross-over design was used during circuit training in physical education, with a repeated-measures factor (music vs. no music) and a between-subjects factor (self-selected vs. provided music). Affective and social variables were assessed pre- and post-lesson; intrinsic motivation was measured post-lesson. The sample comprised 161 students (M age = 10.85, SD age = 0.84). No significant Condition × Time interactions were found for affective valence, arousal, positive or negative emotions, or social connectedness (all ps ≥.230), indicating that pre–post changes did not differ between music and no-music lessons. However, significant main effects of condition indicated higher overall arousal, F (1, 419.00) = 15.43, p <.001, η p ² =.036, higher positive emotions, F (1, 413.14) = 14.60, p <.001, η p ² =.034, greater social connectedness, F (1, 395.76) = 5.19, p =.023, η p ² =.013, and lower negative emotions, F (1, 414.94) = 4.05, p =.045, η p ² =.010, in music compared with no-music lessons. Music was also associated with greater intrinsic motivation, reflected in higher interest/enjoyment, F (1, 136.69) = 7.59, p =.007, η p ² =.053, and perceived competence, F (1, 133.85) = 12.19, p <.001, η p ² =.083. Music selection did not significantly moderate affective, social, or motivational outcomes (all ps ≥.236). These findings highlight music as a feasible didactic tool in PE to enhance students’ arousal and intrinsic motivation and to promote more positive lesson experiences, independent of whether the music is self-selected or teacher-provided.

Physical education (PE) plays a central pedagogical role in shaping students' affective, social, and motivational relationships with physical activity, with the overarching aim of fostering positive movement experiences and supporting lifelong engagement in sport and exercise (Ennis, 2017; Kirk, 2010), thereby acknowledging the holistic perspective on PE (see Carcamo-Oyarzun et al., 2025). Beyond the development of motor skills, PE is increasingly conceptualized as a learning environment in which affective experiences, social belonging, and motivation constitute key pedagogical outcomes (Bailey et al., 2009; Spittle and Byrne, 2009; Standage et al., 2012). Consequently, identifying instructional strategies that enhance students' enjoyment, engagement, and perceived competence during PE lessons remains a central concern within sport pedagogy research. Within this pedagogical context, there is growing support for the intentional integration of music into PE lessons (Reppa, 2007). Music represents a socially accepted, non-invasive, and readily accessible form of environmental modulation that can meaningfully influence students' experiences and behavior during physical activity (Schwartz et al., 2017). Empirical evidence indicates that music in PE settings can increase physical activity levels (e.g., step counts; Brewer et al., 2016; Chen, 2025) and enhance intrinsic motivation, enjoyment, and perceived competence (e.g., Cools et al., 2023). Despite this growing body of evidence, the social functions of music in PE remain comparatively underexplored. In particular, it is still unclear whether the pedagogical impact of music differs depending on whether students listen to self-selected or externally provided music. Allowing students to select music themselves may represent an autonomy-supportive instructional strategy with implications for motivation and social connectedness, yet empirical evidence in PE contexts remains limited. The present study addresses this gap by examining the effects of music use in fitness education–based PE lessons (i.e., circuit training), comparing self-selected and externally provided music, on students' affective experiences, perceived social connectedness, and intrinsic motivation (i.e., interest/enjoyment, perceived competence, effort/importance, tension, and perceived autonomy).

Theories: music on affect, motivation, and connectedness

In the context of physical activity and exercise, several conceptual models explain how music influences psycho-bio-social factors and behavior. A foundational framework by Karageorghis et al. (1999) identifies four hierarchical modulators of affective and physical responses to music: rhythm response (most influential), musicality, cultural impact, and association (e.g., with sports or media). These dimensions shape how music enhances movement and emotional engagement.

Building on this, Lauria (2023) outlines underlying emotional mechanisms such as emotional contagion—mirroring the music's affective tone—and entrainment, where physiological functions such as breathing, heart rate, and bodily movement synchronize with rhythm. These processes can induce approach-oriented tendencies and increase motivation, linking emotional responses to behavioral outcomes.

Another comprehensive approach is provided by Clark et al. (2016) whose meta-theoretical model integrating therapeutic, exercise, and auditory-motor perspectives. They describe how music activates subcortical and cortical pathways, producing affective and arousal effects that are hypothesized to facilitate positive behavioral outcomes such as participation and adherence. Notably, their model includes cognitive effects: music can distract from fatigue and discomfort (DeNora, 2000; Terry and Karageorghis, 2011), altering time perception and bodily awareness—particularly effective during low to moderate intensity exercise (Lind et al., 2009).

While these theories offer valuable insights into mechanisms primarily rooted in psychophysiological responses to music listening, they make only minimal—if any—reference to the social dimensions of musical effects. This conceptual gap is addressed by the Musical Bonding Model proposed by Boer et al. (2011). According to this model, musical taste functions as a social signal for underlying values (Boer et al., 2011; Frith, 1981; Rentfrow and Gosling, 2003). Based on the similarity–attraction hypothesis, shared musical preferences imply shared values, which in turn foster social closeness and interpersonal attraction. Put more simply: “We all enjoy hip-hop music during PE class—this means we are all cool, which makes us feel connected.” Music thus plays a prosocial role by fostering social cohesion and a stronger sense of connectedness, especially during adolescence and in contexts of relationship formation (Boer et al., 2011). This perspective is further supported by Social Identity Theory (Tajfel and Turner, 1979), which posits that individuals derive part of their self-concept from membership in social groups. Within this framework, music could serve as a marker of group identity (see also Rentfrow, 2012), reinforcing the assumptions of the Musical Bonding Model.

In sum, these frameworks collectively highlight music's potential to enhance affective experience, motivation, and social connectedness in exercise settings.

Empirical evidence: music on affect, social connectedness, and motivation in PE

PE is widely understood as a pedagogical setting that shapes not only students' motor competence but also their affective experiences, sense of belonging, and motivation—outcomes that are central to fostering positive relationships with physical activity and supporting lifelong engagement in sport and exercise (Bailey et al., 2009; Ennis, 2017; Kirk, 2010; Standage et al., 2012). In this context, music has been proposed as a promising and practically feasible didactic tool in PE, and recent work has begun to examine its impact on students' experiences and learning-related outcomes (Reppa, 2007).

Within this pedagogical context, a growing body of research has examined the effects of music in PE settings, with most studies reporting positive outcomes for student enjoyment. Increased enjoyment has been observed among children (Deutsch and Hetland, 2012), middle school students (Ha and Wong, 2002), and adolescents (Cools et al., 2023), although findings are not entirely consistent, as one quasi-experimental study with fifth graders reported no significant effects of music on enjoyment (Chen, 2025). Overall, however, existing evidence suggests that music can enhance enjoyment in PE.

Similar patterns have been reported for student motivation. For example, Digelidis et al. (2014) showed that listening to music during PE lessons significantly increased lesson satisfaction and intrinsic motivation among adolescents, while Deutsch and Hetland (2012), using a within-subjects design with different music tempo conditions, found higher motivation in both music conditions compared to a no-music control. Together, these findings indicate that music may support core pedagogical goals of PE by fostering more motivating and enjoyable learning environments.

Beyond affective and motivational outcomes, music may also shape students' social experiences during physical activity, yet empirical evidence within PE remains scarce. This gap is pedagogically important because feelings of belonging and peer acceptance influence students' willingness to participate, take risks, and remain engaged in movement-based learning activities (Hastie et al., 2011; Rikard and Banville, 2006). Evidence from related domains indicates that social closeness can increase when individuals move in synchrony (Hove and Risen, 2009) and when they engage in shared musical activities such as singing or dancing, which may strengthen group cohesion (Tarr et al., 2014). Complementing these findings, the Musical Bonding Model (Boer et al., 2011) suggests that music can facilitate interpersonal connection by serving as a social signal of shared values and identities—especially when musical preferences are experienced as personally meaningful. Accordingly, the social and motivational impact of music in PE may depend not only on whether music is present, but also on the degree to which students identify with the music, for instance when they are given opportunities to select it themselves. However, systematic investigations of these social processes in PE, and particularly the pedagogical implications of self-selected versus externally provided music, remain limited.

Self-selected vs. provided music

From a sport pedagogical perspective, the distinction between self-selected and externally provided music can be understood as an instructional decision that may differentially support students' affective, motivational, and social learning experiences in PE. The notion that self-selected music may exert stronger social-emotional and motivational effects is consistent with the Musical Bonding Model (Boer et al., 2011) and with broader motivational theories. In particular, Self-Determination Theory (Deci and Ryan, 2000) posits that autonomy-supportive instructional practices enhance students' positive affect and intrinsic motivation. Allowing students to choose the music used during PE lessons may therefore represent a pedagogical strategy to support autonomy (Domville et al., 2019), potentially resulting in more favorable affective and motivational responses (see also Cools et al., 2023; Digelidis et al., 2014).

From a social learning perspective, music selection may also have implications for students' sense of belonging. When music choice is negotiated or shared within a class, it may function as a social signal of shared values and group identity, thereby fostering social connectedness in line with the Musical Bonding Model (Boer et al., 2011). Moreover, self-selected music is more likely to activate positive personal associations, which may intensify affective and emotional responses during physical activity, as proposed in the theoretical framework by Karageorghis et al. (1999).

Empirically, however, evidence for differential effects of self-selected versus externally provided music in PE remains limited. Deutsch and Hetland (2012) reported no significant differences in enjoyment or motivation between teacher-selected and self-selected music conditions, and Digelidis et al. (2014) likewise found no differences in motivation or lesson satisfaction among 200 adolescents when music was self-selected versus provided. Taken together, these findings suggest that while theoretical and pedagogical arguments point to potential benefits of self-selected music, the existing empirical evidence in PE is inconclusive. This highlights the need for further research that systematically examines music choice as a pedagogical variable and its effects on affective, motivational, and social experiences in PE settings (see also Priest and Karageorghis, 2008).

The present study

Against this pedagogical background, the present study examines music as a didactic tool in fitness-oriented PE lessons to support students' affective, social, and motivational experiences. This focus is particularly relevant within the context of the Berlin PE curriculum for students in grades 5 and 6, which is guided by the official framework curriculum for sport education across grades 1–10 in the federal states of Berlin and Brandenburg. The curriculum adopts a holistic educational perspective, positioning PE not merely as a setting for the development of motor skills, but as a learning environment that contributes to students' psychosocial, emotional, and cognitive development through meaningful movement experiences. In grades 5 and 6, PE lessons are designed to offer a broad and varied spectrum of movement activities, including athletics games, fitness-oriented tasks, movement sequences, and cooperative challenges. These activities aim to promote physical literacy, enjoyment of movement, social interaction, and reflective judgment, while simultaneously supporting health-promoting behaviors and long-term engagement in sport and physical activity. Importantly, the curriculum emphasizes participation, cooperation, and personal development during a critical developmental phase in early adolescence, while granting teachers pedagogical flexibility in the selection and implementation of instructional approaches and learning environments. As such, PE in Berlin provides an appropriate and authentic context for examining instructional strategies—such as the use of music—that may enhance students' motivation, emotional experiences, and sense of social connectedness during movement-based learning.

Building on this curricular and pedagogical foundation, the present study employed a clustered cross-over design with a repeated-measures factor (music vs. no music) and a between-subjects factor (self-selected vs. externally provided music). We investigated whether integrating music into a fitness-based PE lesson (circuit training) enhances students' affective experiences, perceived social connectedness, and intrinsic motivation compared to lessons without music. In addition, we examined whether allowing students to select the music themselves—conceptualized as a potentially autonomy-supportive instructional strategy—yields stronger effects than externally provided music. By assessing pre–post changes within a single PE lesson and accounting for the clustered classroom structure, the present study addresses methodological limitations of prior PE research and contributes to sport pedagogy by clarifying whether and how music can be used to support motivational and social learning goals in PE.

Based on theoretical assumptions and previous findings, we first hypothesize that playing music during a PE class will positively influence affective (i.e., increased valence, increased arousal, enhanced positive emotions, and reduced negative emotions), motivational (i.e., increased intrinsic motivation), and social parameters (i.e., greater connectedness) compared to having no music at all. Second, we hypothesize that self-selected music, as opposed to externally provided music, will lead to even stronger effects: greater increases in positive affect, positive emotions, and social connectedness; stronger reductions in negative emotions; and more pronounced boosts in intrinsic motivation and social connectedness.

Participants

An a priori power analysis was conducted using GPower 3.1 to estimate the required sample size for detecting the primary effect of condition (music vs. no music). Because GPower does not allow for direct power estimation for linear mixed-effects models, the analysis was based on an F -test approximation for repeated-measures designs, focusing on small-to-moderate effects consistent with previous research on music in physical activity and PE contexts (e.g., Cools et al., 2023; Terry et al., 2020). Assuming a small-to-moderate effect size (f =.20), an alpha level of.05, and a desired power of.80, the analysis indicated a required sample size of N = 182 participants. This estimate was considered conservative, as the mixed-effects models used in the present study account for within-subject dependencies and class-level clustering, which can increase statistical efficiency relative to traditional ANOVA-based approaches. Due to organizational constraints in the school setting (e.g., class availability and absenteeism), the final sample comprised N = 161 students. This sample size was sufficient to detect small-to-moderate main effects, but power may have been limited for detecting smaller moderation effects involving music selection. Importantly, the within-subject crossover design increases statistical efficiency relative to purely between-subject designs, partially compensating for the reduced sample size.

The study was conducted with eight classes; four 5th and four 6th grade classes and a total of 161 students (90 males, 71 females; M age = 10.85, SD age = 0.84) from on school in Berlin, Germany. On average, the students received a PE grade of 1.59 (SD = 0.64) and reported engaging in sports outside of school 3.39 times per week (SD = 1.4). Additionally, 60.2% of the participating students were active members of a sports club.

The study was carried out in accordance with the Helsinki Declaration of 1975. Informed consent was obtained from both the parents and the students. Since the data collection was conducted during the practical semester that university students were required to complete at the Humboldt-Universität zu Berlin, no separate ethics approval from the Berlin Senate and ethical committee was required. This exemption was based on an official decision by the Berlin Senate.

Measures

Questionnaires

Demographic data

Prior to the relevant physical education sessions, participants completed a brief demographic questionnaire during their regular class teacher period. They were asked to provide information on their age, gender, frequency of sports participation outside of school, membership in a sports club, and the grade they received in physical education on their most recent report card.

Affect

The Self-Assessment Manikins (SAM; Bradley and Lang, 1994) were used to detect participants' state of perceived valence and arousal. This non-verbal self-report instrument displayed five abstract manikins and a corresponding nine-point Likert scale, on which participants rated their current emotional state from 1 = unpleasant to 9 = pleasant, and from 1 = low arousal to 9 = high arousal.

Emotions

The German version of the Sport Emotion Questionnaire (SEQ-d; Wetzel et al., 2020) was used to assess the subjective emotional state. The SEQ-d comprised 13 items loading on three dimensions: negative emotions (5 items; e.g., annoyed, disappointed; α =.815), positive emotions (4 items; e.g., happy, content; α =.911), and tension (4 items). Each item represented an adjective (e.g., disappointed, pleased, nervous). Participants were asked to indicate how they felt at that moment on a five-point Likert scale ranging from 0 (not at all) to 4 (extremely). The subscale tension could be interpreted as both positive and negative, which has been shown to lead to ambivalent scores (see Knöbel et al., 2024). Therefore, the tension subscale was not given to participants. Mean scores were calculated for each subscale. Higher mean values indicate a stronger expression of emotions at that moment.

Social connectedness

Self-Determination Theory (Deci and Ryan, 2000) identifies connectedness as a key psychological need, which in this study was assessed using perceived closeness measured by the Inclusion of Others in the Self Scale (IOS; Gächter et al., 2015). This scale used a visual representation of two overlapping circles, where greater overlap (i.e., a larger intersection) was associated with higher perceived closeness. Participants had seven circle arrangements to choose from, ranging from no overlap to nearly complete overlap. Based on the question, “In this moment, how close do you feel to your class?” participants were asked to select the circle arrangement that best represented their relationship.

Intrinsic motivation

Intrinsic motivation was assessed using the Intrinsic Motivation Inventory adapted for PE (IMI-PE; Cocca et al., 2022). The following subscales were included: interest/enjoyment (3 items; e.g., “I really enjoyed PE.”; α =.860); perceived competence (4 items; e.g., “I was quite talented in physical education.”; α =.753); effort/importance (3 items; e.g., “It was important to me to perform well in PE class.”; α =.378); and tension (4 items; e.g., “I felt under pressure in PE class.”; α =.451). Additionally, perceived autonomy was measured using the “freedom of choice” subscale (7 items; e.g., “I participated in physical education because I wanted to.”; α =.613) from the Self-Determination Theory homepage (2024). All items were rated on a 5-point Likert scale (1 = not at all true to 5 = very true). Reverse-coded items were recoded, and mean scores calculated for each subscale; higher scores indicated stronger endorsement of the respective construct. Due to low internal consistency, especially for effort/importance, tension, and autonomy, McDonald's Omega was calculated (Dunn et al., 2014), but yielded similarly poor results. As a result, only the reliable subscales—interest/enjoyment and perceived competence—were retained for further analyses to avoid misleading interpretations (Field, 2024).

Organization of PE lessons – circuit training

The PE lessons included a structured circuit training session in which students rotated through various exercise stations. The circuit consisted of five stations and was implemented in two different versions (Latin square randomized to balance out the order between the conditions) in order to reduce potential boredom (version A: wall sit, plank, leg movements in supine position, lateral jumps, and Russian twist; version B: hip lifts, side planks, leg movements in supine position, step-ups on a bench, and heel taps). Each exercise was presented with differentiated variations to accommodate different fitness levels—for example, performing leg movements in supine position with or without a ball. A start signal marked the beginning of each interval, with time cues provided at 30 and 20 s remaining. The final 10 s were counted down aloud, and an end signal indicated the conclusion of the interval. All exercises were demonstrated once with correct execution, during which the teacher pointed out potential postural errors. During the exercise intervals, students were allowed to adjust the difficulty level of each exercise if they perceived a given variation as too easy or too difficult.

The circuit training consisted of three rounds of five stations. Students exercised for 60 s at each station, followed by approximately 30 s to transition to the next station. If needed, an additional 5–10 s were provided to allow all students to reach the next station. Thus, the scheduled circuit duration was approximately 22.5 min, with minor variations depending on transition times.

Music playlists

For the externally provided music condition, a standardized playlist was created based on popular workout compilations on Spotify. Only the songs with a tempo of at least 120 beats per minute (BPM) were included, as music at this fast tempo has been shown to enhance arousal and increase motivation (Cools et al., 2023).

In the self-selected music condition, students were asked to complete a short form approximately one week prior to data collection, in which they listed up to three preferred songs along with the corresponding artist names. This activity took approximately 10 min and was conducted in a separate classroom. The aim was to compile a class-specific playlist of approximately 45 min to accompany the PE lesson. All songs provided by the students were included and played during the lesson. Full playlists for each class, as well as the standardized playlist including BPM information, are provided in the Supplementary Material. No BPM criterion was applied to these playlists. Mean tempo and BPM ranges were 117.7 BPM (63–173 BPM) for class 5a, 116.0 BPM (89–165 BPM) for class 5b, 125.6 BPM (85–161 BPM) for class 6a, and 124.7 BPM (94–159 BPM) for class 6c.

In both conditions, songs were played in randomized order to ensure a balanced and minimally biased presentation of the musical stimuli. Music was played throughout the entire PE session, including the introductory phase. During instructional periods, the volume was reduced to ensure that verbal explanations remained clearly audible.

Procedure

Prior to data collection (November–December 2024), after consultation with school administration, teachers agreed to implement the study within their PE classes. Parents were informed via letter and asked to provide active consent for their child's participation. In addition, students received a general explanation of the study's purpose and were asked to provide written consent as well.

Each class participated in two PE lessons—one with music and one without—administered in randomized order on separate days. During the music condition, students either listened to self-selected or externally provided music: four classes experienced the self-selected music condition, and four completed the provided music condition. Participants were randomly assigned to these conditions, with equal representation from 5th and 6th grade students to ensure balanced group distribution.

Each PE session followed a standardized structure. At the beginning of each session, students completed questionnaires assessing their affective state (SAM), emotional experience (SEQ), and perceived connectedness to classmates (IOS). This was followed by a circuit training task that was identical across all conditions with regard to timing, organization, duration of exercise and rest intervals, and overall format. At the end of the session, students again completed the SAM, SEQ, and IOS, and additionally filled out the motivation questionnaire (IMI) to assess their motivational experience during the activity.

Data analysis

IBM SPSS Statistics 29 was used for all statistical analyses. All variables were screened for outliers (± 3 SD) and checked for normality. To account for the clustered cross-over design and repeated measurements, data were analyzed using linear mixed-effects models. This approach was chosen to appropriately model the hierarchical data structure, with repeated measurements nested within students and students nested within classes.

For affective and social outcomes (affective valence, arousal, positive emotions, negative emotions, and social connectedness), separate mixed-effects models were specified including condition (with music vs. without music) and time (pre vs. post) as within-subject fixed effects, music selection (self-selected vs. externally provided) as a between-class fixed effect, and their interactions. Random intercepts were included for classes to account for clustering effects and for students nested within classes to account for repeated measurements.

To evaluate the effects of music condition on intrinsic motivation (interest/enjoyment and perceived competence), mixed-effects models were conducted with condition (with vs. without music) as a within-subject fixed effect and music selection (self-selected vs. externally provided) as a between-class fixed effect. Random intercepts for classes and students nested within classes were again specified.

Significant interaction effects were followed up using estimated marginal means and simple effects analyses. Where appropriate, pairwise comparisons were adjusted using Bonferroni corrections. Statistical significance was set at p <.05.

The data were normally distributed, and no outliers were detected. However, not all participants responded to every question, resulting in varying sample sizes across analyses. To maximize the number of responses and the overall data base, all available data were used. Participants were not excluded due to individual missing responses. This approach follows recommendations to handle missing data on an analysis-specific basis, rather than enforcing listwise deletion of cases (see e.g., Newman, 2014).

Descriptive data

Please find the descriptive data in Tables 1, 2.

Descriptive changes in social connectedness across conditions are illustrated in Figure 1.

Music vs. no music

Linear mixed-effects models revealed no significant Condition × Time interactions for affective valence, arousal, perceived connectedness, positive emotions, or negative emotions (all ps ≥.230), indicating that pre–post changes in these outcomes did not differ between music and no-music lessons.

For affective valence, a significant main effect of time emerged, F (1, 388.43) = 5.08, p =.025, η p 2 =.013, 95% CI [.002,.031], reflecting an overall increase from pre to post across conditions. For arousal, significant main effects of condition, F (1, 419.00) = 15.43, p <.001, η p 2 =.036, 95% CI [.015,.064] and time, F (1, 393.12) = 11.06, p <.001, η p 2 =.027, 95% CI [.009,.052] indicated higher overall arousal during music lessons and increased arousal from pre to post. Similarly, perceived connectedness showed significant main effects of condition, F (1, 395.76) = 5.19, p =.023, η p 2 =.013, 95% CI [.002,.031], and time, F (1, 388.52) = 13.29, p <.001, η p 2 =.033, 95% CI [.013,.060] reflecting higher connectedness during music lessons and increases over time. Positive emotions were also higher overall in the music condition, F (1, 413.14) = 14.60, p <.001, η p 2 =.034, 95% CI [.014,.061], whereas negative emotions were lower during music lessons, F (1, 414.94) = 4.05, p =.045, η p 2 =.010, 95% CI [.000,.026].

For intrinsic motivation, significant main effects of condition indicated higher interest/enjoyment, F (1, 136.69) = 7.59, p =.007, η p 2 =.053, 95% CI [.012,.112] and greater perceived competence, F (1, 133.85) = 12.19, p <.001, η p 2 =.083, 95% CI [.030,.155], during PE lessons with music compared to lessons without music.

Self-selected vs. externally Provided Music

Hypothesis 2, which proposed stronger effects for self-selected compared to externally provided music, was not supported. Across all affective and social outcomes, no significant three-way interactions between condition, time, and music selection were observed (valence: F (1, 384.54) = 1.06, p =.304, η p 2 =.003, 95% CI [.000,.016]; arousal: F (1, 527.85) = 0.05, p =.832, η p 2 <.001, 95% CI [.000,.005]; positive emotions: F (1, 533.97) = 1.41, p =.236, η p 2 =.003, 95% CI [.000,.017]; negative emotions: F (1, 534.00) = 0.03, p =.873, η p 2 <.001, 95% CI [.000,.004]; perceived social connectedness: F (1, 530.06) = 0.48, p =.490, η p 2 =.001, 95% CI [.000,.010]). Thus, the effects of music on pre–post changes in affective and social outcomes did not differ between self-selected and externally provided music.

Similarly, for intrinsic motivation, no significant interactions between condition and music selection were found for interest/enjoyment, F (1, 263.96) = 1.29, p =.258, η p 2 =.005, 95% CI [.000,.026], or perceived competence, F (1, 264.40) = 0.91, p =.342, η p 2 =.003, 95% CI [.000,.020], indicating that the motivational benefits of music were independent of whether the music was self-selected or externally provided.

The aim of the present study was to examine the effects of music as an instructional element in physical education (PE) lessons on students' affective, social, and motivational experiences, with a particular focus on whether these effects differ depending on whether the music is self-selected or externally provided. Using a quasi-experimental repeated-measures design and linear mixed-effects models that accounted for the clustered data structure inherent to PE settings (students nested within classes), we examined both general effects of music and the pedagogical relevance of music selection as an instructional decision. Overall, the findings indicate that incorporating music into PE lessons yields robust benefits for students' arousal and intrinsic motivation, while these benefits appear largely independent of whether students select the music themselves or listen to externally provided music.

Across conditions, students showed increases in arousal and perceived social connectedness over the course of the PE lesson, suggesting that participation in PE itself constitutes a socially and physiologically meaningful learning experience. Importantly, lessons that included music further enhanced arousal and intrinsic motivation—reflected in higher interest/enjoyment and perceived competence—compared to lessons without music. In contrast, music selection did not systematically moderate affective, social, or motivational outcomes, indicating that the pedagogical benefits of music were comparable for self-selected and externally provided music.

Effects of music on affective experiences

Consistent with theoretical expectations, music—regardless of whether it was self-selected or externally provided—was associated with higher arousal during PE lessons. This finding aligns with Clark et al.'s (2016) meta-theoretical model, which highlights music's capacity to activate subcortical and cortical pathways involved in arousal regulation. From a pedagogical perspective, this suggests that music can be used as an instructional resource to support students' physiological and psychological activation during PE, even in contexts that are already physically demanding (Lambourne and Tomporowski, 2010).

One plausible explanation is that music functions as a dissociative or attentional support strategy within PE lessons (DeNora, 2000; Terry and Karageorghis, 2011), potentially redirecting students' focus away from discomfort and toward task engagement. If music facilitated higher task involvement or effort, this may have contributed to the observed increases in arousal. However, as perceived exertion was not assessed, this interpretation remains speculative. Future research should therefore include measures of perceived effort, attentional focus, or task engagement to better understand how music operates as an instructional support in PE.

In contrast to arousal, music did not significantly influence affective valence or emotional experiences. This pattern is consistent with recent PE research reporting no effects of music on affective states among fifth-grade students (Chen, 2025). Earlier studies reporting positive emotional effects of music in children (e.g., Deutsch and Hetland, 2012) often relied on composite measures that conflated affect with motivation or performance, limiting their comparability with the present findings. Measurement timing may also have influenced the results. Affect and emotions were assessed at the end of the PE lesson, when students may already have been transitioning cognitively to subsequent classes or experiencing organizational demands (e.g., changing clothes). Such contextual features of PE lessons may attenuate emotional effects and should be considered in future studies using in-task or momentary assessment approaches.

Although no statistically significant emotional effects emerged, descriptive patterns suggested that music may help stabilize positive emotions during physically demanding PE activities. While positive emotions declined in no-music lessons, they remained stable in music lessons—a pattern compatible with the undoing hypothesis (Fredrickson and Levenson, 1998) and recent conceptualizations of music as an emotion-regulation resource in movement contexts (Zajonz and Lautenbach, 2025). These exploratory findings warrant further investigation using designs explicitly tailored to capture emotional regulation processes in PE.

Effects of music on social experiences

Contrary to expectations, music did not significantly influence changes in perceived social connectedness, nor did music selection moderate these effects. Once class-level clustering was accounted for, perceived connectedness increased over time across conditions, suggesting that shared participation in structured PE activities may itself foster social bonding. From a pedagogical standpoint, this underscores the role of PE as a social learning environment in which cooperation, shared effort, and joint task engagement contribute to students' sense of belonging.

These findings suggest that short-term changes in social connectedness during PE may be driven more strongly by instructional structure and collective task involvement than by additional lesson elements such as music. While theoretical models such as the Musical Bonding Model (Boer et al., 2011) emphasize the social potential of shared musical experiences, such effects may be less pronounced in single-lesson PE contexts—particularly when music selection is individualized rather than negotiated collectively as part of instruction.

Effects of music on motivational experiences in PE

In line with pedagogical expectations and prior research, music had a clear positive effect on intrinsic motivation in PE. Students reported higher interest/enjoyment and greater perceived competence during lessons with music compared to lessons without music. These findings align with sport pedagogy research emphasizing motivation as a central instructional outcome in PE and are consistent with evidence from adolescent and adult samples (Digelidis et al., 2014; Dimitriadis et al., 2024).

One possible instructional mechanism is that music facilitates rhythmic alignment between movement and auditory cues (Lauria, 2023), supporting smoother movement execution and enhancing students' perceptions of competence. High-tempo music may have helped structure movement timing during circuit training, thereby contributing to more fluent and enjoyable participation.

Importantly, motivational benefits did not significantly differ by music selection, although self-selected music showed descriptively higher scores. One possible explanation is that class-level song selection may be neither sufficiently individualized to enhance perceived autonomy nor sufficiently collective to strengthen connectedness through shared musical experiences. From a pedagogical perspective, this suggests that music itself—rather than music choice per se —may be sufficient to support motivation in PE, allowing teachers to leverage music without necessarily implementing individualized selection procedures. However, these mechanisms were not directly tested, and the autonomy-related IMI subscale could not be analyzed due to low reliability. Future research should therefore directly assess perceived autonomy and whether music choices are experienced as personally or collectively meaningful.

Strengths, limitations, and future research

A key strength of this study lies in its methodological rigor. By using linear mixed-effects models, we appropriately accounted for the hierarchical data structure and avoided inflated effects that can arise when clustering is ignored. The quasi-experimental crossover design allowed for within-subject comparisons under ecologically valid PE conditions, and standardized instruments were employed to assess affective, social, and motivational outcomes.

Several limitations should be acknowledged. First, the exclusion of some IMI subscales due to low reliability limits conclusions regarding the full motivational profile. Second, the sample was drawn from a single school and restricted to 5th and 6th graders, limiting generalizability. Third, although a self-selected music condition was included, not all students provided sufficient song suggestions, and playlists were partially supplemented, potentially reducing autonomy effects. Similarly, the music conditions were not fully standardized with regard to tempo and playlist duration. Although post-hoc analyses indicated no significant differences in mean BPM between the playlists, the self-selected playlists were not matched to the externally provided playlist with regard to tempo and duration and showed wider BPM ranges. Moreover, music was played throughout the entire PE session rather than being restricted to the standardized circuit training. Thus, variation in musical characteristics and exposure duration cannot be fully separated from the source of music selection. Future studies should standardize music exposure duration and systematically match playlists on relevant musical characteristics such as tempo. Fifth, the absence of measures such as perceived exertion or attentional focus limits insight into underlying mechanisms. Finally, although an a priori power analysis indicated that a slightly larger sample would be desirable, the final sample size of N = 161 students was constrained by practical considerations inherent to school-based research, such as the limited number of available classes and occasional student absenteeism. While this sample size was sufficient to detect small-to-moderate main effects of music, as reflected in the consistent effects on arousal and intrinsic motivation, statistical power may have been limited for detecting smaller moderation effects, particularly those involving music selection at the class level. This limitation is especially relevant given that music selection was manipulated between classes, resulting in a comparatively small number of higher-level units. Nevertheless, the absence of moderation effects was consistent across affective, social, and motivational outcomes. However, given the limited statistical power for detecting smaller moderation effects, these null findings should not be interpreted as evidence for the absence of an effect of music selection. Rather, smaller effects of music selection cannot be ruled out and should be examined in future studies with a larger number of classes.

Future research should incorporate broader motivational measures, include more diverse age groups and school contexts, and assess potential mediators such as perceived exertion, enjoyment, and attentional engagement to better clarify the mechanisms through which music shapes students' experiences in PE. Longitudinal designs would be particularly valuable for examining whether repeated exposure to music-enhanced PE lessons fosters sustained motivation and longer-term physical activity participation beyond the immediate lesson context. In addition, given that music selection was manipulated at the class level, future studies with a larger number of classes and greater variability at the group level are needed to more sensitively test whether subtle moderation effects of self-selected versus externally provided music emerge under different instructional conditions or over longer intervention periods.

In conclusion, the present study demonstrates that music is a powerful and practical tool for enhancing motivational experiences and arousal during PE lessons. Importantly, these benefits were robust across music selection modes, indicating that teachers can effectively use music without needing to individualize music choice. While music did not substantially alter affective or social experiences beyond arousal and motivation, descriptive patterns suggest potential emotion-regulatory functions that merit further study. Overall, the findings support the strategic integration of music into PE as a feasible and impactful approach to fostering engaging and motivating learning environments.