Abstract/Summary

Listening is a crucial language skill for L2 learners, as it is not only one of the most frequently used skills but also plays a vital role in the development of other language skills. However, beyond these benefits, English listening proficiency can also be crucial for Chinese university EFL learners’ academic success, career prospects, and global mobility. For instance, it can impact learners’ exam performance, is essential for meeting job market demands in an increasingly globalised economy, and is a key requirement for studying abroad, where strong listening skills are vital for both academic and daily life. However, Chinese EFL university learners often face numerous challenges in developing their listening proficiency in formal classroom settings, such as limited instructional time, insufficient teaching resources, a lack of diverse teaching methods, and an exam-oriented curriculum. These factors may hinder adequate listening development, leading many learners to experience moderate to high levels of listening anxiety and low self-efficacy. One effective solution is for students to actively engage in informal L2 listening practice outside the classroom. With technological advancements, informal digital learning of English (IDLE) has provided learners with greater access to resources and opportunities. While IDLE has gained increasing research attention, little is known about Chinese EFL university learners’ L2 listening development within informal learning contexts. Moreover, unlike traditional teacher-centered language classrooms, where external regulation is dominant, informal language learning contexts can provide learners with greater autonomy and freedom. In the absence of external regulation, however, learners’ self-regulated learning (SRL) abilities become particularly crucial. While numerous models have been developed to illustrate the mechanisms of SRL, models specifically constructed for L2 listening remain absent. Furthermore, considering the significant impact of motivational and affective factors on listening, regulation of motivation and affect should be incorporated into SRL frameworks. However, few existing models simultaneously address both cognitive (i.e., listening) and motivational/affective regulation. Additionally, SRL, self-efficacy, and listening anxiety have a complex relationship; however, their joint predictive mechanism for listening remains unclear. Therefore, addressing the research gaps above serves as the primary aim of this study. This study employed a mixed-method research design. Based on an analysis of the strengths and limitations of existing SRL models, the study proposed a five-phase, dual-level SRL model targeted L2 listening, referred to as the Self-Regulated L2 Listening Model. To validate the hypothesised structure of the model, a questionnaire was developed based on this theoretical framework, namely the Self-Regulated L2 Listening Questionnaire (SRLLQ). A total of 582 EFL learners from five universities in China were invited to complete the SRLLQ, and 523 valid responses were analysed using Confirmatory Factor Analysis (CFA), which confirmed the five-phase, dual-level structure of the Self-Regulated L2 Listening Model. Additionally, the study proposed two hypothesised joint predictive mechanisms of SRL, self-efficacy, and listening anxiety on L2 listening. To test these mechanisms and explore learners’ L2 listening development in an informal digital learning of English listening (IDLEL) context, another 130 English majors from two of the five universities participated in the IDLEL study. They completed three listening tests (pre, post, and delayed post-test) and two questionnaires (pre, and post-test) measuring their self-regulation, self-efficacy, and listening anxiety. Additionally, they participated in a four-week observational IDLEL study, during which they recorded their IDLEL engagement in their weekly E-logs. The Structural Equation Modeling (SEM) results confirmed the two hypothesised predictive mechanisms: the first mechanism revealed the direct predictive effect of self-efficacy on listening, as well as its indirect effect through SRL and listening anxiety; the second mechanism focused on the direct predictive effect of SRL on listening and its indirect effect, with self-efficacy and listening anxiety being the mediators. Together, these two mechanisms illustrate a positive cycle that facilitates L2 learners’ listening development. Additionally, the multiple predictive pathways of SRL suggest that its influence on listening is not only immediate but also potentially long-term. Moreover, descriptive analysis, thematic analysis, and cluster analysis of participants’ IDLEL E-logs provided insights into the quantity (i.e., frequency and duration of engagement) and quality (i.e., diversity of activities engaged and SRL strategy use) of their IDLEL engagement. Linear Mixed Models (LMMs) were then constructed to reveal the predictive effects of SRL and IDLEL engagement on listening, self-efficacy, and listening anxiety. The results showed that participants’ SRL (post-test) and the duration of their IDLEL activity engagement significantly and positively predicted their listening improvement from pre-test to post-test, but they were not found to have a significant predictive effect on the listening post-test. Additionally, participants’ SRL (post-test) and the duration of their IDLEL engagement significantly and positively predicted their listening self-efficacy at both the pre-test and post-test, while the frequency of IDLEL engagement had a significant negative predictive effect on their listening self-efficacy at both time points. Furthermore, participants’ SRL (post-test) was found to significantly negatively predict their listening anxiety at both time points. Finally, the moderation analysis showed that SRL did not moderate the relationship between IDLEL engagement and listening and self-efficacy, indicating that IDLEL engagement may be universally beneficial to learners, regardless of their SRL abilities.