Extracurricular music and visual arts activities are related to academic performance improvement in school-aged children

The study investigates whether participation in extracurricular music and visual arts activities (arts ECA) relates to general academic performance (ACGP) among Japanese junior high school students and explores potential transfer mechanisms. In Japan, all students receive formal instruction in music and visual arts and may additionally join after-school arts clubs (e.g., brass band/choir, visual arts). Prior research often shows positive associations between music involvement and academic outcomes, sometimes after controlling for SES, though findings are mixed and often correlational. Evidence for visual arts is more controversial, with critiques about lack of controls for baseline ability and SES. Drawing on transfer theory, the authors test two hypotheses: (1) arts education is associated with general academic performance; (2) effects are mediated by improvements in related artistic skills/knowledge (i.e., music and visual arts grades). They conducted a 2-year longitudinal survey, modeling direct effects of arts ECA on ACGP and indirect effects via music (MUGP) and visual arts (VAGP) scores, while controlling for baseline art scores, sex, SES (household income, maternal education), and learning habits. Given prior links between sports ECA and academics, a secondary analysis excluded sports participants.

Music education research has frequently reported positive links between music participation and academic performance, including studies showing associations in national samples after adjusting for SES and family background. Some correlational studies find null results, yet quasi-experimental work suggests music training can elevate IQ and executive functions, which are related to academic achievement. Visual arts education findings are mixed: some report higher SAT or test performance among students engaged in visual arts or Visual Thinking Strategies, but criticisms highlight correlational designs and inadequate controls (baseline intellect, SES). Transfer theory (e.g., Winner et al.) posits that arts education may influence non-arts outcomes through (a) simple transfer (direct, modest effects) and (b) mediated transfer via development of artistic skills, with additional hypotheses on integrated curricula (e.g., STEAM) and real-life outcomes. The present study explicitly tests both simple and mediated transfer hypotheses for music and visual arts within a longitudinal framework, addressing prior limitations by controlling SES and baseline performance.

Design: Two-year longitudinal study following a cohort from the end of 7th grade (first year of junior high) to the end of 9th grade (third year) in Japan. Measures were collected at the end of each school year for academic and arts grades; SES and ECA participation were assessed in the first year (and ECA also in third year). Participants: 946 7th graders recruited from 6 public junior high schools in/near Sapporo; after attrition and exclusions (missing follow-up, missing GP/ECA, missing SES, and students who quit arts ECA), final N=488 (259 male, 229 female). Ages ~12–13 at baseline. A secondary analysis excluded students in extracurricular sports, yielding N=147 (male n=41; female n=106). Variables: - ACGP: latent general academic performance defined by grade points (GPs) in five core subjects (Japanese, Social Studies, Mathematics, Science, English) at each time (first and third years). - MUGP and VAGP: music and visual arts GPs (1–5 scale) at first and third years per national curriculum assessments (rubric-based three-perspective ratings aggregated to 5-point GP). - ECA indicators: music ECA=1 for brass band/choir participants (N=55); visual arts ECA=1 for visual arts club participants (N=23). Other cultural ECA coded separately; sports ECA identified for sensitivity analyses. - SES: household income (5-category from <2M to >8M yen) and maternal education (from junior high to bachelor’s). - Learning habits: latent factor from self-reported study duration on weekdays and weekends at first and third years. Analysis: Structural equation modeling (SEM) with full-information maximum likelihood (R, lavaan). ACGP (first and third year) modeled as latent factors from five subject GPs; learning habits as latent; third-year ACGP, MUGP, and VAGP regressed on first-year counterparts and ECA variables. Control variables (sex, SES, learning habits) included; model refinement excluded non-significant controls (except main ECA/SES paths) and added covariances guided by modification indices (>3.84). Primary model fit: χ²(163, N=488)=485.92, p<0.001; CFI=0.95; GFI=0.99; AGFI=0.98; RMSEA=0.06 (90% CI 0.06–0.07). Secondary model (excluding sports ECA): χ²(175, N=147)=299.08, p<0.001; CFI=0.94; GFI=0.98; AGFI=0.97; RMSEA=0.07 (90% CI 0.06–0.08, p=0.011). Mediation structure tested direct effects of arts ECA on ACGP and indirect effects via first- and third-year art GPs (indirect 1 and indirect 2). An inverse mediation (ACGP mediating ECA→art GPs) showed no significant indirect effects.

- Primary SEM (N=488): • Music ECA: indirect effect via first-year MUGP on first-year ACGP positive (indirect 1: b*=0.055, p=0.004). Direct effect of music ECA on ACGP negative (b*=-0.058, p=0.022). Total effect on first-year ACGP not positive (b*=-0.072, p=0.071). Music ECA positively predicted third-year MUGP (b*=0.194, p<0.001). Indirect effect on third-year ACGP via first- and third-year MUGP positive (indirect 2: b*=0.044, p<0.001). Total effect on third-year ACGP not significant (b*=-0.064, p=0.096). • Visual arts ECA: indirect effect via first-year VAGP on first-year ACGP positive (indirect 1: b*=0.037, p=0.008). Direct effect on ACGP not significant; total effect on first-year ACGP not significant (b*=0.027, p=0.467). Visual arts ECA positively predicted third-year VAGP (b*=0.135, p=0.001). Indirect effect on third-year ACGP via first- and third-year VAGP positive (indirect 2: b=0.028, p=0.003). Total effect on third-year ACGP marginal (b*=0.064, p=0.071). • Model fit acceptable: χ²(163)=485.92; CFI=0.95; GFI=0.99; AGFI=0.98; RMSEA=0.06. - Secondary SEM excluding sports ECA (N=147): • Music ECA: indirect 1 b*=0.102, p=0.008; indirect 2 b*=0.105, p=0.001; total effect on ACGP not significant (b*=0.055, p=0.466). • Visual arts ECA: indirect 1 not significant (b*=0.039, p=0.188); indirect 2 positive but not significant (b*=0.026, p=0.096); total effect on ACGP significant (b=0.168, p=0.016). - Overall pattern: Participation in music and visual arts ECA is associated with higher art subject GPs and, through those improvements, with gains in general academic performance over two years. Direct effects on ACGP are not consistently positive; for music, direct effects were negative, potentially offsetting indirect benefits in total effects. Mediation by art GPs was robust; inverse mediation (ACGP mediating ECA→art GPs) was not supported.

Findings support a mediated transfer from arts ECA to general academic performance: students in music or visual arts clubs improved their art subject grades, and these improvements were associated with gains in overall academic performance across two years, even after controlling for SES, sex, and learning habits. This aligns with transfer theory emphasizing development of domain-specific skills that facilitate broader cognitive and motivational benefits. The negative direct effect of music ECA on ACGP suggests that intensive early engagement (e.g., high practice demands in brass band/choir) may transiently compete with study time, reducing academic performance initially; however, as music skills develop, the indirect benefits via improved MUGP contribute positively to ACGP over time. For visual arts, direct effects on ACGP were not significant, but indirect effects via VAGP were positive, and total effects became significant when comparing arts ECA participants to peers excluding sports ECA, indicating context-dependent salience of the transfer effect. The results underscore that the benefits to academics arise through the outcomes of arts learning rather than mere participation, and they add longitudinal, SES-controlled evidence to a literature often criticized for cross-sectional, uncontrolled designs.

This longitudinal study demonstrates that extracurricular music and visual arts participation relates to improvements in general academic performance, with effects operating primarily through gains in corresponding art subject performance. While direct effects on academics are weak or negative (music) in the short term, the mediated transfer over two years suggests small but meaningful benefits of sustained arts engagement. These findings bolster the case for maintaining and valuing arts education in schools. Future research should employ experimental or quasi-experimental designs to infer causality and unpack mechanisms (e.g., cognitive, motivational, social pathways), include broader contextual comparisons across cultures and school systems, and examine specific pedagogical components of arts ECA that most effectively support academic transfer.

- Causality cannot be inferred: observational, correlational design with self-selected club participation; potential selection effects remain despite controls. - Activity specificity unknown: the study did not detail instructional content or practices within music/visual arts ECA that drive improvements. - Generalizability constraints: effects may vary by cultural, regional, or institutional contexts (e.g., availability/structure of ECA, time commitment). - Sample imbalance: relatively few students in arts ECA led to unbalanced group sizes; results differed across comparison sets (e.g., excluding sports ECA). - Measurement limitations: SES measured at baseline only; academic performance based on school GPs rather than standardized tests.