Return
The dual role of digital literacy in EFL writing: Unmasking its direct and self-efficacy-mediated effects on anxiety among Chinese tertiary learners☆
W
W
DOI:10.1016/j.actpsy.2026.106567.png)
Abstract
En 中文
Focusing on the psychological mechanisms underlying EFL writing in the context of digital transformation, the present study addresses the lack of a comprehensive model integrating digital literacy (DL), EFL writing self-efficacy (EFLWSE), and EFL writing anxiety (EFLWA). To investigate their direct and mediated relationships during the writing process, this study employs structural equation modeling (SEM), informed by social cognitive theory and communicative writing theory. Data were collected from 603 Chinese EFL majors using validated scales. Results revealed that DL positively predicted EFLWSE (/3 = 0.286, p < .001), which in turn negatively predicted EFLWA (/3 =-0.403, p < .001). Notably, DL exhibited a significant positive direct effect on EFLWA (/3 = 0.195, p < .001)-a relationship masked in bivariate correlations (r = 0.051, p = .210) but uncovered via SEM. Mediation analysis confirmed that EFLWSE partially mediated the DL-EFLWA relationship (indirect effect =-0.115, 95% CI [-0.181,-0.065], p < .001), with a suppression effect: the positive direct effect of DL on EFLWA was offset by the negative indirect effect, resulting in a non-significant total effect (estimate = 0.079, p = .378). These findings validate the dual role of DL in EFL writing-bolstering EFLWSE to indirectly reduce anxiety while directly inducing anxiety via technostress or overreliance-by integrating two complementary theories. In practice, the specific effect sizes and suppression mechanism provide empirically grounded guidance for designing targeted DL training and balanced digital tool integration in EFL writing instruction to foster self-efficacy while mitigating anxiety.
Keywords:
Digital literacy
Writing self-efficacy
Writing anxiety
Structural equation modeling
EFL writing
Journal
A
IF:
2.7
Papers:
1.1K
Citations:
8.0K
