Research on written corrective feedback in second language learning has advanced, addressing uncertainties about its effectiveness. The debate now centres on the most effective way to give feedback. However, controversy persists, and research on unfocused feedback remains scarce. This quasi-experimental study examines the effects of unfocused written corrective feedback—direct, indirect, and metalinguistic—on the complexity, accuracy, and fluency of 139 English-as-a-second-language learners’ writing at an English-medium instruction university in the United Arab Emirates. It also investigates the moderating effects of individual-difference variables—language aptitude, attitudes towards written corrective feedback, and second-language proficiency—on feedback efficacy. Three experimental groups and one control group wrote essays (pre-test), revised their drafts (post-test), and composed essays on new topics (delayed post-test) over fourteen weeks. MANCOVA results revealed that participants who received direct and metalinguistic feedback showed reduced fluency compared with those in the indirect feedback and control groups. Lexical diversity was significantly lower in the indirect feedback group than in the direct and metalinguistic groups. Unfocused feedback did not yield significant gains on new writing tasks, and none of the individual-difference variables significantly moderated the effects of feedback on complexity, accuracy, and fluency measures. These findings highlight differential short-term effects of unfocused written corrective feedback on specific writing dimensions, while suggesting limited transfer to new writing tasks and limited moderating influence of learner-level variables. The results provide useful information for practitioners, who could adopt a more blended approach between focused and unfocused feedback and increase the number of treatment sessions.
Authorial voice is an integral component of literary writing where interpretation and analysis play an important role. Based on Hyland (2008), this study focuses on functional lexical bundles, including text, participant, and referential-oriented bundles, to analyze samples of undergraduate students’ writings in literature in an L1 Arabic EFL context. This article aims to investigate the extent to which students’ written essays in literature used lexical bundles as a linguistic resource to assert their individual voice and stylistic identity. The study is significant as it contributes to an understanding of how lexical bundle usage shapes authorial voice among undergraduates, particularly in an L1 Arabic EFL context. The present corpus comprises 157 undergraduate literary essays from a public university in Oman. The corpus is over 90, 000 words. Analysis of frequency, variation, and functional roles is done using Ant Conc 4.0 software. The study adopts a corpus-based quantitative analysis complemented by qualitative contextual analysis by integrating quantitative analysis with textual interpretation to assess learners’ communicative and argumentative competence in academic literary discourse. Findings suggest that students tend to use research-oriented bundles of place and time, with minimal usage of participant bundles, which are key markers of their authorial voice. The study further recommends a targeted intervention through bundle-usage teaching to clarify voice, agency, and authorship in literary analysis.
Listening instruction in many English as a Foreign Language (EFL) contexts continues to rely on globally produced audio materials that are linguistically accurate but culturally distant from learners’ lived experiences. Contextualized pedagogy is widely advocated, yet teachers often lack the resources to produce high-quality localized audio. This study investigates how generative artificial intelligence (AI) supports the development of contextualized listening materials and influences students’ listening experiences. A qualitative multi-site case study was conducted in eight secondary schools across geographically diverse areas of West Kalimantan, Indonesia. English teachers and 32 students participated in the design and classroom implementation of AI-supported listening materials. Data were collected through pre- and post-development teacher interviews, classroom reflections, and post-implementation student interviews. The data were analyzed thematically and compared across cases. The findings show that generative AI functioned not merely as a productivity tool but as a form of pedagogical infrastructure. Through iterative prompt refinement and contextual calibration, teachers transformed local knowledge into pedagogically appropriate listening materials. Students reported improved comprehension, reduced cognitive load, greater engagement, and stronger connections with materials reflecting familiar environments. The study contributes to AI-in-education research by demonstrating how generative AI can strengthen teacher agency and support contextualized listening pedagogy in resource-constrained settings. It highlights the potential of AI to bridge technological feasibility and locally grounded language instruction, contributing to more relevant, inclusive, and responsive EFL education. The study recommends that teacher professional development and reform-oriented artificial-intelligence adoption prioritize prompt-mediated pedagogical reasoning and locally grounded material design.
This study investigates the effect of structured multidimensional teacher support on the writing quality of Saudi English as a foreign language learners and explores how they perceive the support provided during L2 writing instruction. The study provides empirical evidence on the role of structured multidimensional teacher support in improving EFL learners’ writing quality, a topic that has received limited attention in intervention-based research. A quasi-experimental pretest–posttest design was employed with 81 Saudi female undergraduate students enrolled in an academic writing course. Students in the experimental group received structured multidimensional teacher support, whereas those in the control group followed regular writing instruction. Writing quality was assessed using an analytic writing rubric, and learners’ perceptions were measured through a questionnaire. The findings revealed significant improvements in the experimental group’s overall writing quality as well as in content, organization, vocabulary, and language use. However, after controlling for pretest differences, organization was the only component for which the experimental group demonstrated a statistically significant advantage over the control group. Participants also reported highly positive perceptions of teacher support, particularly informational and appraisal support. The findings highlight the potential value of structured teacher support in second language writing instruction. The study recommends incorporating structured multidimensional teacher support into L2 writing contexts and encourages future research to include different learner populations, proficiency levels, writing genres, and longer intervention periods.
This study examines similarities and differences in linguistic features and determines the linguistic features most commonly used in English narrative writing by Grade 6 students enrolled in an English Program and those in a Thai Program at a demonstration school in Bangkok. This comparative quantitative descriptive study provides insights into how different learning programs may relate to students’ linguistic performance in writing, and it helps identify the areas for improvement in linguistics and writing. Various studies have examined the English writing of Thai EFL students; however, research on a comparative analysis of linguistic features in EFL narrative writing remains limited. Three linguistic features are highlighted: lexical, syntactic, and discourse. Forty-nine participants, selected through convenience sampling, produced five narrative texts of about 100 words each. Data were gathered using a narrative worksheet for writing samples; AntConc to tabulate lexical density and diversity; and Coh-Metrix to calculate syntactic complexity, coherence, and lexical cohesion. Descriptive statistics, including percentages and means, were used in this study. The results showed that English Program students wrote more syntactically complex sentences with denser information, possibly due to greater English exposure in their classes. Thai Program students demonstrated better performance in repetition and semantic continuity. Both groups demonstrated similar difficulty with overall coherence. The results have important implications: different language exposure may contribute to variation in linguistic features in narrative writing, and highlight the areas where extra instructional support is needed.
