Xinhua Zhu

Drawing from the established human‐AI collaboration framework (Mosqueira‐Rey et al., 2023), this study investigated how L2 teachers at a research‐intensive university collaborated with Generative AI chatbots to provide feedback on tertiary‐level written assignments over 9 weeks. Data from thirteen teachers included 278 interactions with GenAI, 876 feedback instances, and reflective accounts of their experiences. Findings revealed three Teacher‐GenAI collaboration modes— embedding (human‐led), copilot (interactive), and agent (GenAI‐led)—with a MANOVA indicating that rubric dimensions significantly influenced their frequency of use. Post hoc comparisons indicated that teachers used the agent and copilot modes more frequently in the “language use” dimension. Conversely, the embedding mode was used most frequently in the “content” and “overall evaluation” dimensions. Although the a gent mode reduced teachers' workload in providing feedback on mistakes at the micro level, the embedding mode demanded more effort from teachers to give feedback at the macro level of students' assignments. Guided by Ertmer's framework (1999) on the barriers to educational technology integration, we found that teachers tried to claim more control over human‐AI collaboration by adopting embedding and copilot modes for four concerns: GenAI agents' misalignment with task requirements, inconsistency with external expectations, inadequacy in addressing students' emotional needs, and cautious attitudes toward technology.

Under Hong Kong’s “biliterate and trilingual” language policy, immigrant students often face considerable challenges in reading Chinese. While parental involvement and children’s psychological factors are known to play critical roles in children’s reading development, there remains a lack of research focusing on the factors’ interplay and their effects on second language (L2) Chinese reading, specifically within the unique sociolinguistic context of Hong Kong. To address this gap, this study investigated the influence of parental reading attitudes on reading self-beliefs of first-grade L2 Chinese students and their reading performance. Path analysis results revealed that parental reading attitudes indirectly influenced first-grade students’ L2 Chinese reading performance through reading self-concept and self-efficacy, with self-concept mediating this effect. Self-efficacy had a stronger direct impact on reading performance, while self-concept indirectly contributed via self-efficacy. These findings highlight the role of positive parental attitudes and self-beliefs in enhancing L2 Chinese reading. The implications of the findings are briefly discussed, highlighting their relevance to educational practice and future research.

The emergence of generative artificial intelligence (Gen-AI) tools is fundamentally reshaping the landscape of education. This randomized controlled trial, conducted over an eight-week period, investigates the potential of employing ChatGPT to devise personalized writing program aimed at fostering a growth mindset among second language (L2) learners. A sample of 117 undergraduate students from a Chinese university participated in this study and were randomly allocated to either the experimental ( N = 59) or control groups ( N = 58). Both groups completed two reading-to-write tasks. While the control group engaged in traditional writing tasks, the experimental group undertook personalized writing tasks. Specifically, the reading material was themed around the beliefs of growth mindset, and the detailed content generated by ChatGPT was tailored to students’ individual beliefs and L2 learning experiences. The findings reveal that the personalized writing program effectively bolstered students’ growth mindset in L2 learning. Furthermore, students generally expressed positive attitudes towards the personalized writing program and endorsed the application of Gen-AI in similar learning endeavors in the future. This study introduces an innovative approach to cultivate a growth mindset in L2 learners within authentic learning environment. Besides, it demonstrates the application of Gen-AI in delivering personalized learning experiences for students, providing sample procedures for the generation of prompts to obtain personalized reading materials.

Adaptability refers to a long-term, self-regulating process that enables learners to respond effectively to situational uncertainty and developmental challenges. However, research on adaptability in writing remains relatively limited. Existing studies suggest that adaptability in writing can be supported by cognitive resources like growth mindset (GM), emotional resources like grit, and behaviours resources like strategy use (SU). Yet, little is known about how these resources jointly support writing performance. The present study examined the interrelationships among these three resources from 783 Chinese fourth graders in the Greater Bay Area. Structural equation modelling (SEM) revealed that, overall, grit and SU jointly mediated the association between GM and writing performance. Multi-group SEM analyses further showed region-specific patterns. For Guangdong students, GM played a more prominent role than grit, whereas for Hong Kong students, grit emerged as the stronger contributor. These findings together enrich our understanding of how cognitive, emotional, and behavioural resources support adaptability in writing, thereby contributing to the theoretical account of writing adaptability. Drawing on differences in pedagogical practices between Guangdong and Hong Kong, we further discuss how learning environment may shape adaptive learning mechanisms. Implications are provided for instructional practices to foster students’ adaptive capacities in writing.

Multimodal texts, which integrate elements such as images, tables, and videos, require more strategic navigation and psychological adaptation than single-modal texts. While existing research has acknowledged the importance of emotions, self-efficacy, and strategy use in paper-based reading, how these factors interact in more complex multimodal reading contexts remains largely unexplored, particularly from a person-centered perspective. This study addresses this gap by examining the relationships between self-efficacy, strategy use, and multimodal reading performance in young learners with varied emotional profiles. A total of 915 primary students (mean age = 9.95) completed a one-hour multimodal reading task in Chinese, followed by a questionnaire assessing self-efficacy, strategy use (metacognitive, cognitive, and supportive), and emotions. Latent profile analysis identified three emotional profiles among young multimodal readers: positive, moderate, and negative. Students in the positive profile demonstrated the highest levels of reading performance, self-efficacy, and strategy use. In contrast, students in the moderate and negative profiles showed comparable performance, despite the negative group reporting more frequent use of supportive strategies. Multigroup path analysis revealed that although the impact of self-efficacy on strategy use remained stable across emotional profiles, the effectiveness of strategy use on multimodal reading performance varied. Specifically, metacognitive strategies were most effective for the positive group, supportive strategies benefited the moderate group, and no strategies significantly improved performance for the negative group. These findings underscore the stability of self-efficacy and the critical role of emotions in influencing the effectiveness of strategy use in multimodal reading.