Students’ Perceptions and Digital Literacy Self-Efficacy Toward AI-Based Writing Feedback Systems in English Language Learning

Abstract

This study examined students’ perceptions and digital literacy self-efficacy toward AI-based writing feedback systems in an English language learning context at a technology-oriented higher education institution. The increasing use of automated writing tools, such as grammar checkers and AI writing assistants, has encouraged students to rely on machine-generated feedback during early revision stages. However, their ability to interpret and apply AI suggestions varies depending on their digital literacy and evaluative skills. An explanatory sequential mixed-method design was employed, beginning with a quantitative survey followed by qualitative interviews. Seventy-two students completed the questionnaire, and ten participated in semi-structured interviews. Quantitative results indicated generally positive perceptions of AI feedback systems (M = 3.89, SD = 0.54), particularly in ease of use and initial revision support. Digital literacy self-efficacy was moderately high (M = 3.74, SD = 0.49), although confidence in distinguishing between accurate and inaccurate recommendations was lower. A weak positive correlation (r = 0.32) was found between perception and self-efficacy. Interview findings supported the survey results and highlighted three themes: AI as an initial correction tool, the need for contextual interpretation, and cautious reliance on AI-generated suggestions. Overall, the study emphasises the importance of integrating AI literacy and critical evaluation skills into writing instruction to ensure that AI-based feedback functions effectively as a supportive, rather than substitutive, tool in academic writing.

Ähnliche Arbeiten