Wei He, Shuaiqi Jia, Jinying Wu, Minyan Duan, Xiancheng Zeng
EFA and CFA supported a differentiated first-order structure with good approximate fit (CFI = 0.943, TLI = 0.933, RMSEA = 0.065, SRMR = 0.053), satisfactory composite reliability, and preserved discriminant validity. SEM identified a partially mediated hierarchical structure consistent with the theoretical architecture: physical-cognitive organization was strongly associated with external coherence (β = 0.82), which was in turn the dominant proximal correlate of both physical (β = -0.65) and mental disharmonies (β = -0.68). Bootstrapped mediation indicated robust indirect effects from structural resources to dysregulation via integration pathways, strongest through enacted congruence, while energetic organization retained additional direct regulatory effects. The model explained substantial variance in dysregulation outcomes (R 2 = 0.71-0.72), with acceptable multicollinearity (all VIFs < 3) and stable estimates across split-sample validation. Network analysis of disharmonies revealed a densely interconnected, stable structure (CS-coefficient: strength = 0.52, expected influence = 0.67) with identifiable bridge nodes linking somatic and cognitive-existential indicators, consistent with conditional interdependence rather than reflective outputs of a single latent factor. Convergent validity analyses (N = 76) showed the four SDi LV2 composites correlated substantially and in predicted directions with the WHO-5 Wellbeing Index, the DASS-21, and the DERS, with all 24 primary correlations significant after Benjamini-Hochberg correction (|r| range = 0.39-0.82); the strongest associations fell on the most theoretically proximal benchmarks (wellbeing, depression), and facet-level analyses showed selective convergence (notably DERS Clarity, five facets at |r| ≥ 0.70) alongside near-zero correlations for theoretically distal pairs. Hierarchical regressions showed that integration and disharmonies blocks contributed robustly to predicting external scales, with the emotional resources facet adding substantial incremental variance to emotion-regulation prediction (ΔR 2 = 0.078, p < 0.001).
Beliefs about the nature of athletic ability may shape motivation and engagement in sport, yet their measurement has received limited examination in Chinese populations. This research adapted and evaluated the Chinese version of the Conceptions of the Nature of Athletic Ability Questionnaire-2 (CNAAQ-2-C) across three studies. In Study 1, ordinal exploratory factor analysis with 592 university students supported four factors, Stable, Gift, Improvement, and Learning, χ2(24) = 90.76, CFI = 0.987, TLI = 0.964, RMSEA = 0.069, and SRMR = 0.019. Study 2 replicated this structure in an independent sample of 282 university students; the correlated four-factor model showed good fit, χ2(48) = 84.89, CFI = 0.979, TLI = 0.971, RMSEA = 0.052, and SRMR = 0.040, with standardized loadings from 0.645 to 0.932. Study 3 examined measurement invariance using the complete 12-item model. Configural, loading, and threshold invariance were supported across gender. Across university, primary school, and junior high school samples, configural and loading invariance were supported, whereas evidence for full threshold invariance was mixed. Overall, the CNAAQ-2-C provides a useful four-factor measure for group-level research, although mean comparisons across educational stages require caution.