This study investigates how instruction involving different configurations of explicit information (EI) on contrastiveness (the presence or absence of L1–L2 similarities and differences), practice type, and guided induction (GI) may shape L2 English article learning ( a/an , the , Ø). During an 8-week online intervention, Japanese learners of English as a foreign language ( N = 219) received training under one of four instructional conditions: (1) non-contrastive instruction (rule-based L2–EI + L2-only practice), (2) contrastive instruction (contrastive EI + dual-language practice), (3) contrastive instruction with GI (learner-guided inference from contrastive examples), and (4) an active control receiving matched non-contrastive instruction on prepositions. The primary outcome measure was a timed online grammaticality judgement and correction task (GJCT), administered at pre-test, immediate post-test, and delayed post-test intervals. Mixed-effects modelling showed that all experimental groups outperformed the control, with the contrastive conditions yielding somewhat more favourable outcomes, although the differences among the article-focused conditions remained limited in magnitude. In-treatment performance was also positively, albeit weakly, associated with subsequent GJCT scores. These findings suggest how instructional designs informed by cross-linguistic differences may facilitate L2 article learning in contexts where the L1 lacks an article system.
This article examines how rural boarding school (RBS) English teachers’ agency and identity are enacted through English teaching practices under constraint. Drawing on a qualitative multiple case study of 13 English teachers in three RBSs in northwestern China, the study analyses focus group data, classroom observations, field notes, reflective journals and participant information forms. The findings show how limited family support, fragile learning continuity and perceived institutional and parental risk were intensified in English by high-stakes assessment, restricted exposure and practice and cumulative gaps in earlier learning. Across the three cases, differentiated pedagogical investment took the form of intensified support, concentrated attention, redirection, coercive deterrence or withdrawal. The article proposes selective pedagogical agency as a situated, teacher-level elaboration of educational triage. It refers to the identity-mediated process through which teachers allocated limited time, effort, classroom authority and emotional energy as they interpreted learner need, considered whether further effort could support progress, and weighed emotional demands, moral responsibility and professional risk. Coercive deterrence is analysed as an ethically problematic attempt to force participation rather than as acceptable pedagogy, while resigned permissiveness denotes protective withdrawal when further intervention was judged ineffective, risky or emotionally unsustainable. It argues that rural English teacher identity is not formed outside classroom practice but enacted through repeated decisions about whom to push, whom to support, when to withdraw and how to manage moral and professional risk. This conceptualisation contributes to language teaching and teacher agency research by showing how agency may take the form of selective pedagogical investment when English teachers work under limited support, fragile learning continuity and professional risk. It also offers implications for professional support specific to English teaching, realistic curriculum and assessment expectations and institutional protection for English teachers working in disadvantaged rural contexts.
Task-based language teaching (TBLT) has attracted growing interest as an effective teaching approach grounded in multiple second language acquisition theories. However, previous research has documented that many teachers struggle to implement TBLT despite a foundational understanding of its definitions and principles, highlighting the need for ongoing support. Although teacher training is crucial to successful TBLT implementation, research examining the relationship between TBLT teacher training and subsequent application remains limited. Moreover, TBLT training studies have mainly been conducted in English-as-a-second-language and English-as-a-foreign-language contexts, with less attention paid to international practicum settings, where teacher-learners (TLs) learn to teach in culturally and linguistically unfamiliar contexts. Most studies in international practicums have focused on participants’ development of cultural awareness, while how such experiences enhance teaching practices has been underexplored. To address these gaps, this multiple-case study adopts a bottom-up approach to teacher training to support continuing teacher development. It documents TLs’ growth in TBLT implementation in relation to their perceptions during a two-month practicum in Thailand. Analysis of TLs’ classroom practices and interview responses revealed (a) divergent developmental trajectories, with task-based methodological principles emerging more frequently in their teaching, and (b) positive attitudes toward observation-based feedback provided during debriefing sessions. The study highlights the value of tailoring training to TLs’ needs and levels of expertise, thereby offering insights into essential elements of effective teacher education.
This study investigates the effectiveness of artificial intelligence (AI) use on English learner engagement and conducts moderator analysis to identify the variables that exert a moderating effect on AI’s impact. A multi-database search was conducted to locate 846 records, and 17 primary studies yielding 36 effect sizes were included in this meta-analysis after deduplication, screening and full-text eligibility assessment. Aggregated results show that AI use has a large average impact on English learner engagement ( g = 0.955, 95% CI [0.720, 1.191]). Given the substantial between-study heterogeneity, this pooled estimate represents an average effect across the included studies. Moderator analyses demonstrate that the estimated effect of AI use on English learner engagement varies significantly according to the roles of AI and learners’ major. The results provide researchers, instructors and administrators with the effectiveness of AI in enhancing engagement in English as a second language and a foreign language learning contexts.
Oral corrective feedback (OCF) has been frequently used to address spoken errors, yet the extent to which it improves second language (L2) speaking accuracy remains inconclusive. Previous meta-analyses have blended oral and written measures and are thus unable to answer this question. To narrow this gap, we examined the effects of OCF on grammatical accuracy in L2 speaking through a three-level meta-analysis of 55 primary studies (292 effect sizes, N = 3,352). The results revealed a medium-to-large overall effect ( g = 0.92, 95% CI [0.69, 1.15], p < .001). Moderator analyses provided quantitative evidence for the potential benefits of encouraging learner self-repair and using multiple feedback moves. Notably, the effect of OCF declined significantly over time, a pattern that contrasts with previous meta-analytic reviews. In addition, oral production tasks yielded significantly larger effects than elicited imitation, suggesting that effect sizes vary with testing instruments even within the same modality. These findings were discussed in light of the Noticing Hypothesis, the Skill Acquisition Theory, and the Transfer-Appropriate Processing Hypothesis. The study concludes with pedagogical and methodological recommendations, enriching our understanding of OCF treatment in L2 speaking instruction.
