Recent studies have highlighted the potential of generative artificial intelligence, such as ChatGPT, to address challenges in providing accurate and pedagogically relevant feedback. However, empirical evidence on how prompt engineering shapes feedback quality remains limited. This study examined how zero-shot, few-shot and chain-of-thought prompting strategies influenced the accuracy and depth of ChatGPT-generated qualitative feedback on second language (L2) essays. A total of 176 essays from Filipino and Thai learners with intermediate English proficiency were evaluated using ChatGPT-4o under the three prompting strategies. The findings showed that few-shot prompting achieved the highest accuracy, while chain-of-thought prompting produced the most elaborated feedback, particularly in addressing grammatical complexity. Zero-shot prompting lagged in both accuracy and depth, with notable issues in grammatical feedback. Implications for L2 writing instruction, assessment and research are discussed.

This study investigated the feasibility of using artificial intelligence (AI)-generated reformulations as written feedback in second language (L2) writing contexts. A comparative analysis was conducted to assess the degree of lexical resemblance between reformulated texts by human individuals and those by AI. Following Cohen's definition of reformulation, seven first language (L1) English as a foreign language (EFL) university instructors and three types of AI chatbots − ChatGPT-3.5, Claude2 and Bing Chat Creative Mode − reformulated picture description narratives written by five L1 Japanese EFL university students. Utilizing corpus analysis tools, the study examined text lengths, lexical sophistication, lexical diversity and incorporation of new vocabulary and multiword units. The findings revealed that among all the reformulator types, human-generated reformulations most closely resembled the original drafts in text length and vocabulary usage. Furthermore, among the AI systems, Bing Chat Creative Mode was found to be the most comparable to human reformulators in these aspects. In contrast, ChatGPT-3.5 and Claude2, showcased more advanced vocabulary and greater lexical diversity. The study concludes with pedagogical implications.

English-medium instruction (EMI) has increasingly become a key strategy for internationalization in higher education. However, studies on the perceptions of low-proficiency students remain limited, particularly within technological and vocational education contexts. This study examines the perceptions of 288 students from a Taiwanese technological university regarding EMI and the strategic use of the first language (L1) in EMI classrooms. Quantitative analysis revealed generally neutral to negative attitudes toward EMI, significantly influenced by English proficiency, prior EMI experience and gender. Female students and those with greater familiarity with EMI expressed more favourable perceptions. Qualitative data further indicated that lower-proficiency students highly valued the selective use of L1 by both their instructors and peers, viewing it as essential for content comprehension and active participation. Overall, the results underscore the pedagogical value of selective L1 use to accommodate learner diversity and enhance EMI effectiveness. This study offers practical implications for EMI policy design, classroom practice and instructional planning in an English as a foreign language-based technological university setting.

The paper reports a study that investigated engagement with and response to strategy-based instruction (SBI) targeting grammar learning strategies (GLS). Participants were 36 Polish university students majoring in English. The intervention spanned one academic semester and involved eight 30-minute sessions incorporated into regularly scheduled classes. Data were collected by means of post-intervention questionnaires completed immediately after each session. The students indicated the levels of their engagement at the beginning, in the middle, and at the end of each session on a seven-point scale and responded to open-ended questions. Quantitative analysis showed moderate levels of engagement in all sessions but also indicated that engagement tended to decrease as the sessions progressed. Qualitative analysis identified three main themes, each comprising more specific subthemes: (1) the effectiveness of grammar tasks, (2) new methods of learning grammar, and (3) student engagement. Implications for effective SBI targeting GLS are discussed.

In second language (L2) writing classrooms, collaborative pre-writing (CPW) discussions have emerged as a promising strategy to enhance learners’ writing performance. While previous research has shown that CPW can improve the content quality of individual writing, an important question remains: when learners share a common first language (L1), should they collaborate using L1 or L2 to maximize writing benefits? This study investigates the impact of interaction language (L1 vs. L2) during CPW on the complexity, accuracy and fluency (CAF) of L2 learners’ written texts. Forty university students were assigned into two groups: one conducted pre-writing discussions in L1, the other in L2, before composing individual argumentative essays. The groups then switched the discussion language for a second writing task. All essays were assessed for CAF. Results revealed that L1-based collaboration significantly improved syntactic complexity, while no significant differences were found in accuracy or fluency. The findings contribute to a better understanding of how interaction language in CPW influences writing outcomes, with implications for both theory and classroom practice.

The present study investigated the effectiveness of one form of relatively under-researched advance organizer (AO)-guided concept mapping for English as a foreign language (EFL) learners’ listening comprehension development. The possible mediating role of learners’ initial listening proficiency levels was also explored. A total of 163 Chinese college EFL learners with varying listening proficiency levels were assigned to either an experimental group or a control group. Learners in the experimental group received guidance in drawing a concept map based on the listening passage they just heard and learners in the control group answered some comprehension questions with assistance of strategy training after listening to the passage. A one-way analysis of variance (ANOVA) was employed to analyze the influence of guided concept mapping on listening comprehension development. A two-way ANOVA was applied in investigating if the effectiveness of guided concept mapping was modulated by different listening proficiency levels. The results indicated that guided concept mapping led to moderate gains in listening comprehension development. In addition, skilled listeners achieved significantly greater improvements while their less-skilled counterparts did not benefit from the treatment. Possible reasons behind the results and implications of the study were also discussed.