Related Experiment Videos
Self-perceived digital literacy, creative self-efficacy, and self-reported innovative behavior among undergraduate
Tong Wu1, Xudong Fan1, Xinle Zhou1
1School of Nursing, Wannan Medical University, Wuhu, China.
Background:
Nursing studies have separately linked self-perceived digital literacy with self-reported innovative behavior and the clinical practice environment with creative self-efficacy and innovative behavior. A related association pattern has been reported among Chinese medical students, but joint evidence from multi-institutional nursing samples remains limited.
Objective:
To evaluate the measurement structure and examine the three-variable association pattern among self-perceived general digital literacy, creative self-efficacy, and self-reported innovative behavior in nursing students recruited from three institutions.
Methods:
A multi-institutional cross-sectional online survey conducted from September to December 2025 included 502 undergraduate nursing students from three medical higher-education institutions in Anhui Province, China. Item-level confirmatory factor analyses compared five theoretically plausible measurement models, and convergent and discriminant validity were evaluated before PROCESS Model 4 was applied to observed total scores. The primary analysis used 5,000 bootstrap resamples and eight theory-informed covariates.
Results:
Self-perceived general digital literacy was correlated with creative self-efficacy (r = 0.574) and self-reported innovative behavior (r = 0.584), while creative self-efficacy was strongly correlated with self-reported innovative behavior (r = 0.834; all p < 0.001). The theory-consistent higher-order three-construct model showed acceptable fit (χ 2 = 1314.15, df = 426, CFI = 0.935, TLI = 0.929, RMSEA = 0.065, SRMR = 0.068), fitted substantially better than the one-factor model, and significantly outperformed the corresponding model combining creative self-efficacy and innovative behavior (Δχ 2 = 151.76, Δdf = 2, p < 0.001). Their latent correlation was 0.907, HTMT was 0.902 (95% CI, 0.858-0.941), and the Fornell-Larcker criterion was not met. In the adjusted observed-score model, the total coefficient, direct coefficient, and bootstrap product term involving creative self-efficacy were 0.145 (95% CI, 0.127-0.163), 0.040 (95% CI, 0.025-0.055), and 0.105 (95% CI, 0.087-0.124), respectively. The product-term confidence interval excluded zero, and its estimate was materially similar without covariate adjustment. A supplementary post hoc reparameterization based on the observed-score correlations indicated that the data did not uniquely determine the directional ordering.
Conclusion:
In this pooled three-institution nursing-student sample, the three self-report constructs showed a coherent cross-sectional association pattern across measurement-model comparisons and adjusted and unadjusted specifications. This finding connects two previously separate lines of nursing evidence and extends the complete pattern beyond a single-campus nursing sample. Creative self-efficacy was a theoretically relevant correlate within the contemporaneous statistical decomposition, not an established temporal or causal mechanism. Longitudinal studies with objective or observer-rated measures are needed to test temporal ordering and educational or practice-relevant implications.
