Perceived psychological safety, state social connection, and student flourishing in collaborative music activities: a three-wave longitudinal study

The pursuit of excellence in higher music education frequently demands rigorous technical precision, deep discipline, and continuous public evaluation. While collective environments like choirs, orchestras, and chamber ensembles are staple components of university training, new empirical research highlights that these settings do not automatically yield positive psychological outcomes for every student. A comprehensive three-wave longitudinal study tracking 341 university music students across three Chinese universities has revealed that an individual’s perceived psychological safety within a collaborative music group acts as a critical precursor to long-term well-being, mediated through immediate state social connection.

Led by researcher Shihao Zhang, the study investigated whether feeling secure enough to voice concerns, acknowledge errors, and offer diverse artistic interpretations correlates with broader personal flourishing. Rather than examining musical synchrony or collective auditory expression, the research focused strictly on general interpersonal dynamics. The resulting data indicate that while group music-making offers unique spaces for artistic collaboration, student well-being depends heavily on the interpersonal climate established within these environments.

Chronology and Study Design

The empirical investigation utilized a structured three-wave longitudinal survey design to map changes over time and control for baseline metrics. Researchers recruited participants from 23 distinct music-group clusters—including symphonic orchestras, choirs, traditional Chinese instrumental ensembles, chamber groups, contemporary bands, and composition circles.

Data collection occurred across three distinct intervals spaced approximately six weeks apart (averaging 42 days between waves). Out of an initial pool of 378 respondents, 341 participants satisfied all prespecified data-quality parameters and completed the evaluation phases. The retention strategy employed a dynamic branching methodology, allowing researchers to accurately capture shifts in participation status, such as temporary leaves of absence or structural ensemble changes, without introducing traditional monotonic attrition bias.

The 6-week window between waves was deliberately selected to span multiple rehearsal and feedback cycles, providing enough duration for a proximal relational experience to evolve while keeping the overarching activity context stable. Core variables were meticulously assessed using validated scales: Edmondson’s Team Psychological Safety Scale adapted for musical contexts, the UBC State Social Connection Scale, and Diener’s Flourishing Scale.

Core Quantitative Findings

The primary structural equation modeling (SEM) and confirmatory factor analyses yielded robust statistical evidence supporting the hypothesized sequential pathway. Cross-sectional and longitudinal evaluations demonstrated strong internal consistency, with Cronbach’s alpha and McDonald’s omega reliability values consistently exceeding 0.93 across all three survey waves. Furthermore, all core psychological measures achieved scalar measurement invariance over time, validating that the constructs were interpreted equivalently across the assessment periods.

The path analysis revealed that perceived psychological safety measured at Wave 1 (T1) significantly predicted higher state social connection at Wave 2 (T2) ($beta = 0.350, p < 0.001$). In turn, state social connection at T2 strongly predicted student flourishing at Wave 3 (T3) ($beta = 0.274, p < 0.001$). Crucially, the longitudinal indirect effect linking T1 perceived psychological safety to T3 flourishing via T2 social connection was statistically significant ($B = 0.0949, p < 0.001, 95%$ BCa CI $[0.054, 0.142]$), whereas the direct path between T1 safety and T3 flourishing was negligible and statistically non-significant ($beta = -0.004, p = 0.941$).

Additional comparative evaluations between competing structural models—including reverse-ordering and alternative sequence models—demonstrated that the hypothesized developmental pathway fit the data substantially better. However, information criteria presented mixed signals regarding strict directionality: while the Bayesian Information Criterion (BIC) favored the more parsimonious hypothesized model, the Akaike Information Criterion (AIC) slightly favored a reciprocal specification. Consequently, auxiliary evaluations indicated a plausible bidirectional relationship, wherein earlier social connection also significantly predicted subsequent perceptions of psychological safety ($beta = 0.134, p = 0.004$).

Methodological Robustness and Contextual Controls

To ensure the reliability of the findings against potential confounding variables, the research team implemented a series of rigorous sensitivity checks. Because participants were nested within 23 music-group clusters, cluster-robust standard errors were calculated to account for potential group-level dependencies. The resulting standard-error ratios remained remarkably close to 1.0 (ranging from 0.81 to 1.06), indicating minimal within-group dependence.

Furthermore, a Random-Intercept Cross-Lagged Panel Model (RI-CLPM) was deployed to separate stable between-person differences from time-varying within-person fluctuations. At the within-person level, the indirect pathway remained significant ($B = 0.1697, p < 0.001$), reinforcing the interpretation that transient increases in an individual’s psychological safety relative to their own baseline corresponded with heightened subsequent social connection and flourishing.

Demographic covariates such as age, gender, and cumulative years of formal music training showed no significant predictive power over the core constructs. Controlling for survey-timing variations and missing-data patterns via full information maximum likelihood (FIML) and inverse-probability weighting confirmed that attrition did not systematically skew the reported parameters.

Academic and Practical Implications

The implications of this research extend far beyond the theoretical modeling of positive psychology, offering actionable insights for educators, conductors, and institutional administrators within conservatories and university music departments.

Traditional music pedagogy frequently emphasizes absolute technical perfection, discipline, and uncompromising accuracy. While these pedagogical priorities remain vital for artistic development, the study underscores that an unyielding focus on performance quality can inadvertently foster an atmosphere of public comparison, fear of error exposure, and professional isolation. When students perceive that admitting a mistake or questioning a musical interpretation carries social or academic penalties, they often retreat into self-protective silence.

The empirical data suggest that cultivating environments where psychological safety is prioritized can fundamentally alter the social fabric of student ensembles. By encouraging transparent communication, validating constructive inquiries, and framing technical errors as natural components of artistic experimentation, educators may help students bridge the gap between technical isolation and authentic social belonging. Enhanced state social connection subsequently acts as a psychological catalyst, fostering the meaning, engagement, and optimism essential for comprehensive student flourishing.

Ultimately, the research frames collaborative music-making not as a guaranteed generator of well-being, but as a dynamic interpersonal arena. Future investigations within music education are encouraged to examine specific ensemble leadership structures, incorporate direct behavioral observations, and deploy multi-level frameworks to further decode how artistic collaboration impacts mental health and academic resilience.