Evidence remains limited on how self-reported learner autonomy changes when EFL students practise the same language material through immersive VR and desktop activities. This quasi-experimental study compared changes in two existing classes of Omani foundation-programme students during six weeks of vocabulary instruction. Forty pre-intermediate learners participated, with 20 students in each class. One class used immersive VR, while the other completed desktop activities based on the same 20 words and expressions, weekly schedule, spoken-response format, and teacher. Participants completed the 23-item Language Learner Autonomy Scale before and after instruction. A 2 × 2 mixed ANOVA examined the Time × Group interaction for the total score, followed by secondary analyses of the six subscales. The mean total score increased from 71.45 to 90.40 in the VR class and from 69.65 to 73.05 in the desktop class, producing a 15.55-point difference between their mean gains. The Time × Group interaction was statistically significant, F(1, 38) = 208.95, p < .001, partial η² = .846, and the VR class reported larger gains on all six subscales. The subscale pattern connected immersive vocabulary practice with students’ reported motivation, persistence, and management of learning. The study contributes a pretest–posttest active comparison from an underrepresented Omani foundation-programme context and relates immersive task design to students’ perceived responsibility during structured digital practice under closely matched instructional conditions.
The growing prevalence of English Medium Instruction (EMI) in higher education has prompted closer consideration of academic-language development, multilingual participation, linguistic inequality, and the preparation of teachers working in English-medium systems. EAP and related language-support teachers occupy a distinctive position because their responsibilities centre on language and academic communication. The professional capability through which they reason when pedagogical, linguistic, and institutional considerations support competing responses remains insufficiently theorised. This conceptual paper develops multilingual pedagogical reasoning as a framework for language-teacher development in multilingual EMI higher education. Using a problem-driven theory synthesis, the paper engaged selected EMI, EAP, teacher-cognition, critical multilingual, language-management, agency, and professional-identity scholarship iteratively and purposively, assigning a distinct explanatory function to each retained theoretical resource. The synthesis defines multilingual pedagogical reasoning as situated professional judgement through which language-support teachers interpret consequential language-related dilemmas, consider plausible responses, and justify choices under particular pedagogical, linguistic, professional, and institutional conditions. The resulting framework distinguishes reasoned judgement from enactability and presents multilingual practice as one possible response among several. It incorporates evidence-informed reconsideration and positions professional identity as a site where responsibility and legitimate expertise are reconsidered. Four relational propositions and explicit scope conditions make these claims available for empirical examination. The framework offers a theoretically specified basis for professional learning centred on authentic pedagogical dilemmas. It remains provisional and open to qualification, reformulation, or rejection through practitioner appraisal and subsequent empirical investigation.
Research has documented English use in multilingual Mandarin classrooms, but little is known about how beginners in Arabic-dominant universities experience English mediation over time or how those accounts relate to investment-related self-reports. This longitudinal mixed-methods case study examined beginning-to-end changes in five provisional questionnaire item groupings and students’ accounts of English, Arabic and Mandarin during a 14-week beginner Mandarin course at a university in northern Jordan. Data comprised 34 matched questionnaires, repeated interviews with ten students, learner reflections, classroom observations and course documents. Mean scores rose in all five groupings. The largest differences concerned situated access, legitimacy and social support (mean difference = 0.69, p < .001, dz = 0.94) and imagined multilingual and professional futures (mean difference = 0.64, p < .001, dz = 0.88). Qualitative accounts described English as a bridge when it was accessible, served a defined purpose and was followed by Mandarin activity. Gatekeeper accounts concerned a hidden prerequisite, a benchmark based on long-established English ability, and the substitution of English explanation for Mandarin practice. Students also described Mandarin as a prospective specialist resource in an Arabic–English–Mandarin repertoire and reported changes in beginner legitimacy. The cohort-level increase coexisted with stable and declining individual profiles and with reports of both bridge and gatekeeper experiences. The bridge–gatekeeper distinction connects investment-related questions of access, legitimacy and anticipated capital with students’ accounts of how multilingual classroom activity was sequenced in an Arabic-dominant setting. The study assessed questionnaire responses and reported experiences; English and Mandarin proficiency and causal instructional effects were not tested.
Research on Collaborative Online International Learning (COIL) groups interpersonal, cultural, professional and critical reflection as intercultural competence, leaving their evidential basis unclear. This qualitative case study examined forms of reflection among 146 pre-service teachers in Mexico and Chile and educational meanings attributed to a four-month English-mediated South–South COIL project. Data comprised 51 voluntary midpoint essays and four final-panel transcripts from the full cohort. A codebook-based thematic analysis classified non-exclusive relational, descriptive, professional and power-aware reflection and recorded English-use and participation conditions separately. Cultural comparison (67%), interpersonal rapport (61%) and regional or national identification (51%) were most prevalent in the essays. Panel prevalence was highest for interpersonal rapport (51%), cultural comparison (39%), explicit references to structural inequality or unequal authority (37%) and anticipated diversity-related classroom application (36%). Accounts of English referred to anxiety, peer assistance, Spanish clarification and assessed participation. Power-aware passages addressed Anglophone educational authority, social-class inequality, professional devaluation and unequal educational conditions. English was a communicative medium, a participation condition and a subject of power-aware reflection. South–South contact accounts included regional affiliation, unequal global educational relations and future teaching. The classification reserves criticality for explicit references to hierarchy, coloniality, institutional inequality or unequal authority, increasing precision in claims about intercultural reflection. The findings supply Latin American evidence that COIL permits joint examination of language use, intercultural relations, professional intention and inequality. Dataset differences in timing, prompts, format and participation preclude inference about participant-level development or classroom enactment.
No transparent, reproducible synthesis has examined service-learning research in language education published after 2017 while jointly covering foreign-language, ESL/EFL and teacher-education settings. This scoping review mapped recent contexts, participants, research designs, documented outcomes and supporting evidence. Scopus and ERIC searches identified 47 eligible sources: 44 primary empirical reports and 3 reviews. The primary reports represented an estimated 42 distinct studies. Review conduct followed scoping-review guidance, reporting followed PRISMA-ScR, and all final eligibility decisions were made by the author after AI-assisted full-text screening. The primary corpus spanned 14 countries but remained concentrated in United States institutions, English and Spanish target languages, and higher education or pre-service teacher education. Qualitative designs predominated, and nearly all reports included self-reported or perceived evidence. Linguistic, professional, intercultural and civic outcomes commonly co-occurred. Only five reports used assessed measures, two included comparison groups, four followed participants for more than one course or project cycle, and eight collected direct community evidence. Fully online service appeared in three reports, and generative AI was directly investigated in one. The synthesis separates publication counts from underlying studies and outcome prevalence from evidential strength. It establishes that the literature documents several categories of participant learning and experience but provides insufficient evidence of consistent effectiveness, durable change or reciprocal community benefit. Reciprocity and socially mediated learning define service-learning, yet recipient perspectives remain largely absent from its empirical record.
