Introduction

As China’s economy develops, public participation in cultural and artistic activities has grown significantly. Choral singing, as a highly accessible group art form, has seen rapid expansion in China due to its low barriers to entry (e.g., no need for individual instruments) and strong institutional support1. Motivation plays a critical role in choir participation and retention, making it a key research topic for educators and psychologists. According to the China Choral Association Annual Report (2023)2, over 500,000 registered participants engage in university, workplace, and community choirs nationwide, with government-backed initiatives like the National Choral Festivals further driving engagement. Besides, 4.7 million Chinese university students participate in campus choirs annually based on Ministry ofEdcation China (2022) report3. Choirs attract both students and adults seeking to continue singing after graduation, particularly in higher education, where holistic development, including artistic literacy, is increasingly valued. By providing a platform for musical expression, teamwork, and emotional growth, choirs help cultivate aesthetic appreciation and self-identity. Given this context, understanding sustained student interest in choirs is a timely and underexplored research area.

The concept of learning intention (LI) has been studied extensively over the past decades across both academic and practical fields Lent and Brown4. In recent years, music education has become one of the important carriers of aesthetic education5. Individuals can express their emotions and deepen their understanding and perception of musical works through learning to control their breath, master rhythm and pitch accuracy in choir singing, and listen to their own and others’ voices in the process of vocal collaboration6. Even though multiple studies have shown that music can improve students’ performance, and some have evaluated student satisfaction at the end of music courses, there has not been a study conducted to determine college students’ expectations of them before joining a choir and their continuous learning intention after joining7. To bridge the knowledge gap, this study investigates the motivational factors for college students to participate in choir singing and their retention, as well as the relationship between music factors (i.e. the quality of song content and personal music literacy) and psychosocial factors (social group and perceived self-efficacy) in choir participation. In addition, this study also contributes to the current literature by exploring the roles of perceived usefulness and satisfaction. Perceived usefulness reflects that in a choir, students feel the charm and joy of music, thereby enhancing their learning confidence. Moreover, satisfaction is reflected in the effectiveness of the course through students’ own perception.

Comprehensive quality is one of the important criteria for measuring talents today8. Participating in a choir not only enhances students’ musical literacy, but also exercises their organizational skills, leadership abilities, time management, and other aspects9,10 (Hallam 2021)11. Thus, it is crucial to investigate how choir art promotes students’ growth, as these insights can help understand the motivations and influencing factors of college students’ participation in choir clubs and enrich the development of choir art teaching. From the practical point of view, the insights from this study can help better leverage the role of choirs in enhancing students’ overall quality, providing stronger support and guarantees for their comprehensive development. From the theoretical perspective, these findings provide a new perspective on investigating students’ learning intention in choir activities, which remains an underexplored area in the current literature12. This study further promotes the formation of a detailed mechanism, including the relationships between the aforementioned factors.

Theoretical background

Concept of continuous learning intention

Learning intention is defined as the degree to which an individual judges their tendency to engage in a certain learning behavior13. Learning intention is also a state of individual involvement in group activities, which includes explicit behaviors such as whether the individual is present, interacting and communicating with other individuals, as well as emphasizing the individual’s internal psychological characteristics and the ways and degrees of mutual influence between the individual and the group14. The intention to participate in the choir refers to whether college students participate in it, as well as the process of college students exerting subjective initiative to intervene in choir work to achieve club goals and promote self-development15. Research has shown that college students voluntarily choose to participate in choirs based on their own interests, abilities, and needs16. Music itself can enhance sensitivity, including the beauty of music and lyrics, which can be described in religious terms to create a sense of peace and joy. Emotions can affect music participation17. Hearn18 found that by designing, implementing, creating, and producing high-quality choir performances for students, their participation can be enhanced. Chung19 believed that the higher the perceived usefulness of students, the stronger their continuous learning intention. Guo20 proposed that the accumulation of direct experience can enhance learners’ intention to participate in learning activities, Rahimi21 determined that learning interest is intrinsic value and has a positive impact on learning intention.

Theories of expectation confirmation theory

The Expectancy Confirmation Theory (ECT) was first proposed by Oliver in 1980, which suggests that consumers’ intention to continue purchasing products or services is mainly determined by their satisfaction after purchase, and satisfaction is determined by both expectations and confirmation levels. Mortimer et al.,22 extended the theory of expected confirmation, stating that expectations, performance, confirmation, and satisfaction collectively affect sustained purchase intention. Oghuma23 found that expected confirmation, satisfaction, self-efficacy, perceived usefulness, and perceived pleasure all have a direct or indirect positive impact on the intention to continue using. Connor Desai24 supported that perceived usefulness, perceived fun, perceived ease of use, and other factors have an impact on the intention to continue participating. Dai et al.25 proposed that perceived usefulness, perceived service quality, and satisfaction have an impact on the intention of virtual community users to continue participating. Besides, many studies have applied expectancy confirmation theory to the field of learning, such as Mailizar et al.26 posits that satisfaction, confirmation, perceived usefulness, and flow experience have a direct or indirect positive impact on students’ continuous learning intention. Thi et al.,27 proposed that expected confirmation, perceived usefulness, perceived ease of use, course quality, peer effects, and learning desire have a direct or indirect impact on students’ continuous learning intention. In short, it is expected that the confirmation theory plays a crucial role in evaluating students’ continuous learning of the choir. Therefore, this study evaluated both perceived usefulness and satisfaction to reflect the influencing factors of sustained learning among students in actual choirs28.

Theoretical integration: a combined framework for sustained choir participation

This study proposes that these two theories are complementary rather than competing, addressing different but sequential phases of the participation process. The Theory of Learning Intention primarily explains the antecedent motivational drivers that lead individuals to initiate participation. It focuses on pre-engagement factors such as personal interest, perceived value, and self-efficacy beliefs that motivate the decision to join a choir15,16. In contrast, Expectation Confirmation Theory (ECT) elucidates the post-participation evaluative processes that determine whether individuals continue their engagement. It emphasizes how the congruence between pre-participation expectations and actual experiences (confirmation) shapes perceived usefulness and satisfaction, which in turn drive continued intention25,26. This integrated approach allows us to examine not only what brings students to the choir but also what makes them remain, particularly how experiential factors like song quality and social dynamics influence post-participation evaluations.

Hypotheses

Social group

Research on college choir shows that social factors are important predictors of enrollment rates29. The data suggests that the presence of a common language among a group of learners may contribute to this finding30. For example, choirs use specialized, goal-oriented language to share experiences during rehearsals and performances31. The shared experiences during rehearsals and performances have brought these people closer to each other’s social understanding. In large-scale studies, different social groups will have different styles of choir song content6. Other studies have found that the overall perception of social benefits and positive social activities is valuable for choir members and can enhance learners’ self-efficacy32. In this study, we examine how the perceived social environment (Social Group) influences an individual’s belief in their choir-related capabilities (Perceived Self-Efficacy). Hence, the hypotheses formulated as follows.

H1

Social group positively correlates with song’s content quality.

H2

Social group positively correlates with perceived self-efficacy.

Song’s content quality

Song’s content quality greatly affects students’ enthusiasm and participation in the choir33. As a collective form of artistic expression, the choir’s selection of songs not only reflects the choir’s artistic level and aesthetic orientation, but also directly affects the members’ learning interests, emotional investment, and teamwork ability investigated the perceived benefits of choir members and found that aesthetic benefits were rated the highest by including high-quality repertoire, musical style, challenging pieces, improved discipline, and greater artistic appreciation34,35. These factors have contributed to the aesthetic motivation of participating in choirs and enhanced the self-efficacy of college students. Meanwhile, Tu et al.36 stated that choir members with experience in singing advanced songs exhibit higher aesthetic responses compared to those without choir experience. In addition, Deldjoo et al.37 concluded that spirituality is related to artistic creativity and the feeling of separation from society, or conversely, the feeling of close communication with others. In this study, we examine how the perceived quality of the artistic material (Song’s Content Quality) influences an individual’s belief in their choir-related capabilities (Perceived Self-Efficacy). Hence, the following hypotheses are constructed.

H3

Song’s content quality positively correlates with perceived self-efficacy.

H4

Song’s content quality mediates the relationship between social group with perceived self-efficacy.

Perceived self-efficacy

Perceived self-efficacy is defined as an individual’s ability and belief to complete tasks and achieve established goals. Elam et al.,38 emphasized that self-efficacy is an internal reinforcing force. When individuals feel capable of performing a certain behavior, they will feel an inner satisfaction due to their own judgment of the ability, which promotes personal development39,People with high perceived self-efficacy tend to have strong confidence in the tasks they face, while those with low self-efficacy may feel unable to complete them40. The level of self-efficacy affects people’s decision-making41. Thus, perceived self-efficacy plays a crucial role in students’ participation in choir activities. In this study, perceived self-efficacy specifically refers to academic self-efficacy, which can be defined as an individual’s belief in their ability to handle academic tasks. Hence, the following hypotheses are constructed.

H5

Perceived self-efficacy positively correlates with perceived usefulness.

H6

Perceived self-efficacy positively correlates with participation behavior.

H7

Perceived self-efficacy positively correlates with satisfaction.

Perceived usefulness

The concept of perceived usefulness originates from Davis’ TAM model, which refers to the extent to which users perceive that new information technologies can help them perform or work better42. Research has shown that perceived usefulness has a direct effect on learning behavior intention43. Roca demonstrated that enhancing learners’ sense of usefulness can increase their intention to engage in learning behavior44. Fernandez-Herranz et al.45 also showed that the more scholars feel that a certain learning field has a relative advantage, the more they want to participate. Therefore, the following assumptions were established. Hence, the hypotheses are formulated as follows.

H8

Perceived usefulness positively correlates with participation behavior.

H9

Perceived usefulness mediates the relationship between perceived self-efficacy and participation behavior.

Satisfaction

Satisfaction has a significant impact on students’ intention to participate in choir learning46. When students feel higher satisfaction in the choir, they are more likely to maintain a positive learning attitude and strong intention to learn47. The various activities organized by the choir, such as rehearsals, performances, exchanges, etc., the richness, fun, and effectiveness of these activities will directly affect students’ enthusiasm for participation and learning motivation (Musicae Scientiae 2023). Deci believed that students with high self-efficacy are usually able to express their musical ideas and feelings more accurately, thereby improving their learning satisfaction48. Hence, the hypotheses formulated as follows.

H10

Satisfaction positively correlates with participation behavior.

H11

Satisfaction mediates the relationship between perceived self-efficacy and participation behavior.

Personal music literacy

Individual literacy can influence participation in music49. Music can convey specific emotions such as happiness, sadness, and anger through various musical elements, which may include rhythm, melody contours, pronunciation, dynamics, harmony/dissonance, pitch, timbre, and the complexity of harmony, melody, and rhythm50. Therefore, personal emotions towards singing are revealed as an important factor. Students with good music perception ability can have a deeper understanding of the emotional connotations of songs, thus better expressing the emotions of songs during singing51. Meanwhile, students with high personal music quality are often better able to collaborate with team members and complete choir tasks together52. In addition, they are willing to invest more time and energy to enhance their musical abilities, thereby improving their individual and team performance levels. Hence, the following hypotheses are constructed.

H12

Personal music literacy correlates with the relationship between social group and song’s content quality.

H13

Personal music literacy correlates with the relationship between perceived self-efficacy and satisfaction.

H14

Personal music literacy correlates with the relationship between perceived self-efficacy and perceived usefulness.

Conceptual model

The conceptual model is shown as Fig. 1:

Fig. 1
Fig. 1
Full size image

Research model.

Methods

Study design and participants

This study employed a cross-sectional survey design to investigate the factors influencing college students’ continued intention to participate in choir activities. Participants were recruited from Hainan Normal University using a convenience sampling method. Inclusion criteria were: (1) current enrollment as an undergraduate student at the university, (2) participation in at least one choir activity (e.g., rehearsal, performance) within the current academic year, and (3) willingness to provide informed consent. Exclusion criteria included: (1) incomplete survey responses and (2) no prior choir experience.

A sample size justification was performed based on recommendations for Structural Equation Modeling (SEM). A common rule of thumb suggests a minimum participant-to-observed-variable ratio of 10:153. Our measurement model contained 33 observed variables, indicating a required minimum sample of 330. Our final sample of 315 valid responses closely approximates this threshold and is deemed sufficient for model testing, as it exceeds the widely used minimum of 200 for complex SEM models.

Measures

The questionnaire used in this study aims to comprehensively explore the motivation and influencing factors of college students’ participation in choir. The questionnaire covers three key aspects, the first focusing on social groups, which includes experience of problem related social groups and teamwork, as well as social interactions and collective cooperation experiences in choirs. Second, song’s content quality, personal music literacy, perceived self-efficacy was used to evaluate the content quality of songs, examine individuals’ understanding and musical literacy of music, and explore individuals’ self-efficacy in choir performance, that is, their confidence and ability in music expression and performance. Third, this study uses perceived influence and satisfaction, including perceived usefulness, satisfaction、participation behaviour to evaluate students’ perceived usefulness of choir activities, as well as to investigate students’ satisfaction and analyse their participation in choir behaviour.

All constructs were measured using items adapted from established scales in the literature. Responses were recorded on a five-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). The complete English version of the questionnaire is provided in Appendix A. Table 1 summarizes the constructs, their sources, and definitions. All data collected during this study were fully anonymized to protect the confidentiality of the participants. Data were collected between February and June 2024, with a total of 315 valid questionnaires collected from participants recruited at Hainan Normal University. The ethics committee waived the requirement for written (signed) consent. After agreeing to take part in the study (ethical approval details are provided in the Ethical Approval section), participants completed an anonymous online questionnaire distributed via the Wenjuanxing platform. The survey required approximately 10–15 minutes to complete.

Table 1 Construct definitions and measurement.
Full size table

Data analysis plan

Data analysis was conducted using SPSS 26.0 and AMOS 26.0. The analytical procedure followed a two-step approach: (1) confirmatory factor analysis (CFA) to validate the measurement model, and (2) structural equation modeling (SEM) to test the hypothesized relationships. Missing Data: Cases with missing values on key study variables were excluded using listwise deletion, affecting less than 2% of the total sample. Control Variables: Demographic variables (age, gender, grade) were initially included as covariates in the SEM. As their effects were non-significant and did not change the interpretation of the main model paths, they were removed to present a more parsimonious final model. Path Coefficients: In the SEM results, we report unstandardized estimates (B), standard errors (SE), standardized estimates (β), 95% confidence intervals (CIs), and exact p-values. Mediation Analysis: The significance of indirect (mediation) effects was tested using the bootstrap method with 5,000 resamples, generating bias-corrected 95% confidence intervals. Subgroup Comparisons: Differences in participation behavior across demographic groups (age, gender, grade) were examined using one-way ANOVA and independent samples t-tests, with effect sizes (Cohen’s d) reported where applicable.

Results

Basic information analysis of chorus questionnaire

This study conducted an extensive questionnaire survey on college students who participated in the choir, aiming to comprehensively understand their motivation and willingness to continue learning when participating in the choir. As shown in Fig. 2, this study analysed the participation behaviour of college students in different age groups. There are significant differences in the behaviour of students in different age groups when participating in choirs. Students aged 18–20 show a high level of enthusiasm in participating in choir activities, with a significant number of them expressing “agreement” or “strongly agreement” with their participation. Students in the age groups of 21–23 and 24–26 also showed a certain level of positivity in their participation behaviour, but relatively less. Young college students show a higher interest and willingness to participate in choir activities, possibly due to their stronger curiosity and enthusiasm for new things. Second, this study analysed the relationship between gender and participation behaviour. There are also certain differences in the behaviour of men and women in participating in choirs. Women are more active in participating in choirs than men, with a significantly higher number of women expressing “agreement” and “strongly agree” to participate in choirs compared to men. The reason may be that women have higher needs for emotional expression and social interaction, and choir activities perfectly meet these needs.

Fig. 2
Fig. 2
Full size image

Basic analysis results of choir questionnaire.

Finally, this study analysed the relationship between grade level and participation behaviour. There are also significant differences in the behaviour of students from different grades participating in choir. Freshman and sophomore students show high enthusiasm in participating in choir activities, while junior and senior students have relatively less participation. This may be because younger students face less academic pressure and have more time and energy to participate in extracurricular activities, while older students face more academic and employment pressure and have relatively less time to participate in extracurricular activities. The basic analysis of questionnaire data not only reveals the basic characteristics of participants, but also provides rich information for this study, helping to design more targeted choir activities and improve students’ participation and satisfaction. Formal statistical tests confirmed these observations. A one-way ANOVA revealed a significant main effect of grade level on participation behavior, F(3, 311) = 6.52, p < 0.001. Post-hoc tests indicated that freshmen and sophomores reported significantly higher participation than juniors and seniors (p < 0.05). An independent samples t-test showed that female students (M = 4.21, SD = 0.76) reported significantly higher participation intention than male students (M = 3.85, SD = 0.82), t(313) = 3.21, p = 0.001, Cohen’s d = 0.36.

Factor analysis

  1. (1)

    Reliability and validity analysis:

This study conducted factor analysis to evaluate the reliability and effectiveness of the questionnaire. First, the Cronbach’s Alpha coefficients of each item in the questionnaire were calculated. Table 2 shows that the overall Alpha coefficient is 0.926, indicating that the questionnaire has high internal consistency and further verifying the reliability of the questionnaire measurement tool. Second, KMO (Kaiser Meyer Olkin) measure is used to evaluate the adaptability of the samples. The results showed that the overall KMO value was 0.914, far higher than the recommended standard of 0.5, indicating that the sample is very suitable for factor analysis. This confirms that the collected data has good sampling adaptability and can be effectively used for subsequent factor analysis. When conducting factor analysis, the Approx. Chi Square test is used to evaluate the fit of factor structures. According to the analysis results, the approximate chi square value is 5679.956, the degree of freedom is 528, and the p-value is less than 0.001, indicating that the model is statistically significant and supports our proposed factor structure hypothesis. Therefore, the questionnaire has good reliability and effectiveness in this study through the above analyses, providing solid methodological support and data foundation for further exploring the motivation and influencing factors of college students’ participation in choir.

Table 2 Reliability and validity.
Full size table
  1. (2)

    Factor analysis.

The key to factor analysis is to determine the appropriate number of factors to effectively explain the variable relationships and overall variance in the data. From the total variance explanation table, the cumulative percentage of the first few factors is 59.301%, indicating that these factors can already explain most of the variance of the data well. As the number of factors increases, the cumulative percentage continues to rise, but the growth rate gradually slows down. The first 10 factors have already explained approximately 74.326% of the total variance. When deciding how many factors to extract, it is necessary to balance explanatory power and model simplicity. Choosing to extract the top 10 factors is a reasonable decision, as they can effectively capture most of the data variation, and beyond this number, the contribution of newly added factors to the overall interpretation gradually decreases. Therefore, the first 10 factors are sufficient to construct a relatively concise model that can fully explain the data structure. Next, we will further analyze the components and loadings of each factor, explore their relationship with specific variables, and investigate their actual meanings and explanations in the research topic of college students’ participation in choir motivation and influencing factors. This will help this study to gain a deeper understanding of the intrinsic structure behind the factors and their contribution to the research question (Table 3).

Table 3 Results of factor analysis.
Full size table

Confirmatory analysis

  1. (1)

    Model indicator analysis:

Confirmatory factor analysis evaluates the degree of fit between the constructed factor structure model and actual observed data through a series of model fitting indicators. From the table, the fitting indicators include CMIN/DF, NFI, RFI, IFI, TLI, CFI, GFI, and RMSEA. Firstly, the CMIN/DF (Chi Square/degrees of freedom) value is 1.313, which is within the acceptable range, indicating a good fit of the model. Usually, the closer the CMIN/DF value is to 1, the better the fit between the model and the data. Secondly, various comparative fitting indicators such as NFI (Normalized Fit Index), RFI (Relative Fit Index), IFI (Incremental Fit Index), TLI (Tucker Lewis Index), and CFI (Comparative Fit Index) all exceeded the standard of 0.9, indicating a good fit between the model and the data. The high values of these indicators indicate that the model has high accuracy and credibility in interpreting observed data. In addition, the GFI (Generalized Fit Index) value of 0.907 also exceeds the standard of 0.9, indicating that the model can predict the covariance matrix of the observed data well, thus verifying the overall adaptability and accuracy of the model. Finally, the RMSEA (Root Mean Square Error Approximation) is 0.032, which is lower than the commonly accepted standard of 0.08, indicating a good fit between the model and the observed data, with small errors and strong predictive ability of the model. In summary, the model fitting index based on confirmatory factor analysis shows that the established factor structure model fits well with actual observed data. All indicators indicate that the model has high reliability and validity in explaining the relationships between explanatory variables and its interpretability to observed data (Table 4).

Table 4 Model fit of validation factor analysis.
Full size table
  1. (2)

    Convergence validity analysis:

Based on the convergence validity analysis results in Table 5, we can delve into the importance and explanatory power of the Average Variance Explanation (AVE) and Build Reliability (CR) of each variable in the study. Convergence validity analysis is used to evaluate the effectiveness and consistency of measurement tools for latent variables in a measurement model, which helps to confirm the internal structure of the model and the accuracy of the measurement tools. First, according to the AVE values in Table 5, the AVE of each latent variable is at a relatively high level, with values of 85.3, 76.9, 81.2, 79.5, 73.4, and 82.7, respectively. These high AVE values indicate that each variable explains approximately 80% or more of the variance when measuring the latent construct it represents, demonstrating the effectiveness and accuracy of the measurement tool in capturing the concept of latent variables. Specifically, a high AVE value reflects the ability of the measurement tool to be highly correlated with its underlying latent variables, which is crucial for ensuring the reliability and validity of research results. Second, constructing reliability (CR) values is also an important indicator for convergent validity analysis, used to evaluate the internal consistency and reliability of measurement tools. From the CR values in Table 5 (0.92, 0.85, 0.89, 0.88, 0.82, and 0.90, respectively), the CR values of each variable are higher than the commonly accepted standard (usually 0.70), indicating that the measurement tool has good consistency among different variables. This consistency confirms that measurement tools can stably capture the potential constructs represented by each variable, making research results more reliable and accurate. Therefore, in the convergent validity analysis of confirmatory analysis, high AVE and CR values not only confirm the effectiveness and reliability of the measurement model but also provide a solid foundation for subsequent causal relationship analysis and model comparison. These results emphasize the rationality and accuracy of the measurement tools used in the study, which will contribute to further empirical research and the derivation of conclusions.

Table 5 Results of convergence validity analysis.
Full size table

Model path analysis

  1. (1)

    Structural Equation Modeling Path Analysis.

The results of the structural equation model path analysis are presented in Table 6. The standardized path coefficients (β) indicate the strength and direction of the hypothesized relationships. As shown in Table 6, several key hypotheses were supported. Social Group had a significant positive effect on both Song’s Content Quality (H1: β = 0.36, p < 0.001) and Perceived Self-Efficacy (H2: β = 0.41, p < 0.001). Perceived Self-Efficacy, in turn, was a strong predictor of both Perceived Usefulness (H5: β = 0.30, p < 0.001) and Satisfaction (H7: β = 0.44, p < 0.001). Finally, both Perceived Usefulness (H8: β = 0.30, p < 0.001) and Satisfaction (H10: β = 0.41, p < 0.001) significantly predicted Participation Behavior, confirming their roles as proximal determinants of continued intention. Contrary to our expectations, the direct paths from Song’s Content Quality to Perceived Self-Efficacy (H3: β = 0.11, p = 0.167) and from Perceived Self-Efficacy to Participation Behavior (H6: β = 0.03, p = 0.554) were not statistically significant. This suggests that the influence of song quality on self-efficacy, and of self-efficacy on behavior, may be fully mediated by other variables in the model (as tested in the following mediation analysis).

Table 6 Structural equation modeling path analysis results.
Full size table
  1. (2)

    Analysis of mediating effect path in structural equation modeling.

The mediating effects of three important hypotheses (H4, H9, and H11) were explored, which play a key role in understanding the complex mechanisms of college students’ motivation to participate in choirs. According to the experimental results in Table 7, H4 demonstrates the significant mediating effect of song content quality between social groups and perceived self-efficacy. The quality of song content enhances individuals perceived self-efficacy by influencing social group perception, with a significant mediating effect of 0.470 (95% CI = [0.276, 0.667]), indicating how song quality indirectly affects participation behaviour through this pathway. The total effect is 0.930 (95% CI = [0.695, 1.163]), with P values all less than 0.001. Moreover, based on the analysis results of H9 and H11, we found that perceived usefulness plays an important mediating role between perceived self-efficacy and participation behaviour. The indirect effect of H9 is 0.593 (95% CI = [0.406, 0.783]), indicating how perceived self-efficacy affects an individual’s actual behaviour through their perceived usefulness in participating in activities. The indirect effect of H11 hypothesis is 0.847 (95% CI = [0.655, 1.082]), which demonstrates how perceived self-efficacy indirectly affects subsequent participation behaviour through satisfaction after participating in activities. In addition, the combined total effect of H9 and H11 was 1.473 (95% CI = [1.210, 1.727]), further confirming their significant joint and mediating effects on participation behaviour. Through these analyses, not only have we gained a deeper understanding of the important roles of song content quality, perceived usefulness, and satisfaction in influencing college students’ motivation to participate in choirs, but we have also revealed how they interact in individual psychological mechanisms.

Table 7 Bootstrap results for mediation effects (5,000 resamples).
Full size table
  1. (3)

    Analysis of personal music literacy correlates with the relationship between social group and song’s content quality.

The following is a step-by-step analysis of each path (see in Table 8). First, the estimated value of the constant path is 1.5479, the standard error is 0.6513, the critical ratio (C.R.) is 2.3766, and the significance level is 0.0181 (* p < 0.05). This indicates that the influence of constants on the model is significant, suggesting the existence of a fundamental influence in the model, which needs to be carefully considered in model interpretation. Second, for the SG (Social Group) pathway, its estimated value is 0.4215, standard error is 0.1845, critical ratio is 2.2839, and significance level is 0.0231 (* p < 0.05). This means that the influence of social groups on the model is significant, supporting the importance of SG variables in the model. Social groups may play an important role in the quality of music content, which is consistent with research hypotheses. Subsequently, the estimated value of the PML (Personal Music Literacy) pathway was 0.2882, with a standard error of 0.1796, a critical ratio of 1.6045, and a significance level of 0.1096 (p > 0.05). The impact of PML was not statistically significant, but it may still have some indirect effects or moderating effects on other variables in the model, and further exploration of its potential role is needed. Finally, the estimated value of the SG × PML interaction term is − 0.0379, with a standard error of 0.0485, a critical ratio of − 0.7827, and a significance level of 0.4344 (p > 0.05). This indicates that the interaction between social groups and individual music literacy is not statistically significant, suggesting that there may not be a significant interaction effect between these two variables. In summary, in the second part of the structural equation modeling path analysis, we explored in detail the moderating effect of individual music literacy on the relationship between social groups and song content quality. These analysis results not only enhance the understanding of the interrelationships between variables in the model, but also provide guidance and inspiration for further research in the future.

Table 8 The result of personal music literacy correlates with the relationship between social group and song’s content quality.
Full size table
  1. (4)

    Analysis of personal music literacy correlates with the relationship between between perceived self-efficacy and satisfaction.

The following is a step-by-step analysis of each path and moderation effect (see in Table 9). First, the estimated value of the constant path is 2.2820, the standard error is 0.5052, the critical ratio (C.R.) is 4.5167, and the significance level is 0.0000 (***p < 0.001). This indicates that the constant has a highly significant overall impact on the model, indicating that the fundamental influence in the model is very stable and significant, which is crucial for the explanatory power of the overall model.

Table 9 The result of personal music literacy correlates with the relationship between between perceived self-efficacy and satisfaction.
Full size table

Next, the estimated value of the Perceived Self Equity (PS) pathway was − 0.0155, with a standard error of 0.1377, a critical ratio of − 0.1128, and a significance level of 0.9102 (p > 0.05). This indicates that perceived self-efficacy has no significant impact on other variables in the model, and may not play a dominant role in this model. Then, the estimated value of the PML (Personal Music Literacy) pathway was 0.0204, with a standard error of 0.1375, a critical ratio of 0.1480, and a significance level of 0.8824 (p > 0.05). Although the impact of PML was not statistically significant, its potential influence or moderating effect on other variables in the model still needs further exploration. Finally, the estimated value of the interaction term for PS × PML is 0.0973, with a standard error of 0.0362, a critical ratio of 2.6854, and a significance level of 0.0076 (* * p < 0.01). This indicates a significant interactive effect between personal music literacy and perceived self-efficacy. Specifically, when an individual’s level of music literacy is high, the perceived self-efficacy has a more significant impact on satisfaction, which further emphasizes the important role of personal music literacy in regulating the relationships between other variables in the model.

In addition, the conditional effects of regulatory effects further support this finding. When the adjusted variable value increased from 2.0000 to 4.5000, the effect increased from 0.1790 to 0.4222, and the corresponding t-value and significance level (P-value) also significantly increased. The confidence intervals (LLCI and ULCI) further confirmed the stability and significance of this effect. In summary, in the third part of the structural equation modeling path analysis, we explored in detail how personal music literacy regulates the relationship between perceived self-efficacy and satisfaction. These results not only contribute to a deeper understanding of the interactions between model variables, but also provide important theoretical and practical guidance for subsequent research.

  1. (5)

    Analysis of Personal music literacy correlates with the relationship between between perceived self-efficacy and perceived usefulness

Table 10 shows that the estimated value of the constant path is 2.8839, with a standard error of 0.5564, a critical ratio (C.R.) of 5.1835, and a significance level of 0.0000 (* * * p < 0.001). This indicates that the constant has a highly significant overall impact on the model, indicating that the fundamental influence in the model is very stable and significant, and is crucial for the explanatory power of the overall model. Subsequently, the estimated value of the Perceived Self Equity (PS) pathway was − 0.1294, with a standard error of 0.1517, a critical ratio of − 0.8531, and a significance level of 0.3943 (p > 0.05). This indicates that perceived self-efficacy has no significant impact on other variables in the model, and may not play a dominant role in this model.

Table 10 The result of personal music literacy correlates with the relationship between between perceived self-efficacy and perceived usefulness.
Full size table

Then, the estimated value of the PML (Personal Music Literacy) pathway was 0.0589, with a standard error of 0.1514, a critical ratio of 0.3889, and a significance level of 0.6976 (p > 0.05). Although the impact of PML was not statistically significant, its potential influence or moderating effect on other variables in the model still needs further exploration. Finally, the estimated value of the interaction term for PS × PML is 0.0802, with a standard error of 0.0399, a critical ratio of 2.0096, and a significance level of 0.0453 (* p < 0.05). This indicates a significant interactive effect between personal music literacy and perceived self-efficacy. Specifically, when an individual’s level of music literacy is high, the impact of perceived self-efficacy on perceived usefulness is more significant, which further emphasizes the important role of personal music literacy in regulating the relationships between other variables in the model.

Besides, the conditional effects of regulatory effects further support this finding. When the adjusted variable value increased from 2.0000 to 4.5000, the effect increased from 0.0309 to 0.2313, and the corresponding t-value and significance level (p value) also significantly increased. The confidence intervals (LLCI and ULCI) further confirmed the stability and significance of this effect. In summary, in the fourth part of the structural equation modeling path analysis, we explored in detail how personal music literacy regulates the relationship between perceived self-efficacy and perceived usefulness. These results not only contribute to a deeper understanding of the interactions between model variables, but also provide important theoretical and practical guidance for subsequent research.

Discussion

This study analysed the complex relationships between social groups, song content quality, perceived self-efficacy, perceived usefulness, participation behaviour, satisfaction, and personal music literacy through structural equation modeling. As shown in Table 10 and Fig. 3, social groups were found to be significantly positively correlated with song content quality, and song content quality played a mediating role between social groups and perceived self-efficacy. These findings support the H1 and H4 hypotheses. However, the study failed to support the H12 hypothesis, which suggests the relationship between individual musical literacy and the quality of social groups and song content (Table 11).

Fig. 3
Fig. 3
Full size image

Research hypothesis path coefficient.

Table 11 Hypothesis classification.
Full size table

Besides, perceived self-efficacy was found to be significantly positively correlated with perceived usefulness and satisfaction, supporting the H2, H5, and H7 hypotheses. The study did not support the H6 and H8 hypotheses, which suggest a positive correlation between perceived self-efficacy, participation behaviour, and perceived usefulness. Regarding satisfaction, this study supports the mediating role of satisfaction between perceived self-efficacy and participation behaviour (H11 hypothesis), while the mediating role of perceived usefulness between perceived self-efficacy and participation behaviour is also supported (H9 hypothesis). Overall, this study revealed a complex network of relationships between social groups, song content quality, perceived self-efficacy, perceived usefulness, participation behavior, and satisfaction through systematic structural equation modeling analysis. These results not only contribute to a deeper understanding of the interactions between these variables but also provide important theoretical and empirical support for further research in related fields.

Implications

Theoretical implications

This research enriches the current knowledge of choir literature in several ways. First, unlike previous studies that mainly investigated motivational factors for choir participation, this study examined the complex relationship between social groups, song content quality, perceived self-efficacy, perceived usefulness, participation behaviour, and satisfaction. This has not been mostly explored, but it is a behavioural result of students actively participating in choirs55. Second, research has found that social groups have a significant positive impact on the quality of song content, and song content quality plays a mediating role between social groups and perceived self-efficacy. This discovery provides important clues for understanding the psychological mechanisms of individual behaviour under group influence, revealing the potential pathways through which groups influence individual behaviour and attitude formation. Third, perceived self-efficacy has been shown to play an important role in perceived usefulness and satisfaction, reflecting how individuals’ beliefs in their own abilities affect their perception and level of participation in products or services56. This is crucial for understanding and predicting psychological factors in the decision-making process, especially when it comes to the adoption and participation of new products or services. Lastly, the theoretical implications of this study not only deepen our understanding of the interaction between complex social and psychological factors at the theoretical level but also provide a theoretical foundation and guidance for practical applications in related fields57. Future research can further explore and validate these findings to promote further theoretical development and in-depth exploration of empirical research.

Practical implications

The results generated from this research provide some insightful practical implications, which can benefit the students actively participate in the choir. First, the guiding teacher of the choir must ensure the quality of the song content, as popular songs will make the students’ experience interesting and enjoyable (Greer,2020). One way to achieve this goal is to promote engaging songs that immerse students in singing, thereby stimulating their desire to learn. Second, this study combines self-efficacy with choir training, filling the gap in academic research in this field. It analyzes the learning subjects in the choir training process from the perspective of educational psychology, breaking through the traditional model of studying choir training from the teaching process, and enabling choir training to obtain solutions from a new perspective (Maury, 2022). In addition, the practical implications of this study provide important guidance and reference for choirs to better understand students’ psychological and behavioral factors. By deeply understanding and applying research results, students’ performance and satisfaction in choir can be effectively improved, thereby achieving an impact on personal growth, teamwork, artistic cultivation, and mental health of college students58.

Limitations and future research

This study has several limitations that should be acknowledged. First, the generalizability of findings is limited by the sample, which was drawn solely from Hainan Normal University via convenience sampling. Results may not extend to other university populations or cultural contexts. The unique institutional culture and student demographics of a single university limit the broader applicability of our conclusions. Second, the cross-sectional design and reliance on self-reported questionnaire data constrain causal inference and may introduce common method bias. The absence of behavioral or observational data means our findings are based entirely on participant perceptions at a single time point. This mono-method approach also means we lack complementary data sources, such as behavioral observations or instructor assessments, that could provide a more multidimensional understanding of participation dynamics. Third, our theoretical model, though comprehensive, necessarily omits certain contextual factors that may influence choir participation. Variables such as the pedagogical style of the choir conductor, the level of institutional support for arts programs, and specific peer influence mechanisms outside our measured constructs were not examined. Their exclusion means our model may not capture the full complexity of the choir participation ecosystem. Future research should employ multi-site and longitudinal designs to enhance generalizability and establish causality. Incorporating mixed methods and contextual variables would provide a more comprehensive understanding of choir participation dynamics.

Conclusion

Given the requirements of aesthetic education, the participation of college students in choir activities fully demonstrates the value and significance of general music education in universities. This study proposed a comprehensive framework that elucidates the impact of music factors (such as song content quality and personal music literacy) and psychosocial factors (including social groups and perceived self-efficacy) on college students’ participation in choir. This study further emphasizes two behavioural characteristics, namely perceived usefulness, and satisfaction. In sum, this study is considered an additional step in understanding the attitudes and behaviours of college students participating in choirs, providing strategic direction for choir training. This study provides empirical support for university choirs and suggests that future research should further explore the relationship between teacher emotions and student satisfaction, as well as how to strengthen the connection between students and choirs through educational strategies and promote innovation and cultural inheritance in music education.