The book “Academic Writing as if Readers Matter” by L. Cassuto, published by Princeton University Press in 2024, presents academic writing as professional communication by scientists and about science. The topics addressed include the pillars of academic writing which implies doing research and reporting it as a conversation working toward greater knowledge and better understanding. The author explains the main stages of producing a piece of academic writing, from analyzing the title, finding the best sources, note-making and referencing, through to editing the paper, peer review of the work and its publishing. The book aims at providing the reader with tools and techniques helping to communicate science to the public.
Background. The Discussion section is where research findings are reported, interpreted, evaluated, connected to previous work, and turned into claims or implications. It is challenging for novice writers and EFL learners due to its complex rhetorical demands. Previous studies have focused on either rhetorical moves or metadiscourse markers (MDMs) in isolation, but much less is known about how particular markers help writers perform particular moves, and whether these patterns differ across languages and disciplines. Purpose. This study examines how MDMs are distributed across rhetorical moves in the Discussion sections of Chinese and English research articles from soft and hard science. It seeks to identify disciplinary and linguistic differences in move-based marker use and explore how these patterns can inform genre-based academic writing instruction. Method. The corpus included 200 Discussion sections from Scopus Q1 (English) and CNKI core (Chinese) journals. Texts were coded for rhetorical moves using an integrated move framework and for MDMs using Cao’s taxonomy. Frequencies were normalized per 1,000 words. Chi square tests, Cramer’s V, adjusted residuals, and qualitative examples were used to identify significant patterns. Results. Disciplinary differences were move specific. In English articles, soft sciences used more attitude markers when drawing implications (15.17 vs 6.39), while hard sciences relied more on endophoric markers in background information and reporting results (3.88 vs 0.38 and 6.09 vs 1.69). In Chinese articles, soft sciences used more frame markers and boosters when reporting results (4.81 vs 1.53 and 8.05 vs 2.67), whereas hard sciences used more endophoric markers in reporting and commenting on results (6.76 vs 1.29 and 3.70 vs 0.64). The most striking cross-linguistic difference concerns self-mentions. English articles used them significantly more frequently than Chinese articles across several moves in both soft and hard sciences (p < .001, Cramer’s V=0.176-0.507). An Argumentative Progression Pattern also emerged: writers shift from interactive to interactional markers as they develop interpretations, claims, evaluation and implications. Conclusion. The findings demonstrate that MDM use in Discussion sections is shaped by discipline, language, and publication context. They also suggest that genre-based academic writing pedagogy should teach MDM as a move-sensitive resource, especially self-mentions, hedges, attitude markers, and endophoric markers.
Background. Bilingual subtitles can facilitate incidental vocabulary learning by providing simultaneous access to L2 forms and L1 meanings, but they also increase split-attention demands in multimodal processing. In such contexts, textual enhancement may either support noticing by increasing the perceptual salience of target words or add to cognitive load. Although previous research has examined bilingual subtitles, textual enhancement, and individual differences, these factors have rarely been investigated together within a single design. As a result, it remains unclear how textual enhancement operates in bilingual subtitle environments and to what extent its effects depend on learners' working memory, language proficiency, and vocabulary size. Purpose. Given the lack of integrated evidence on how textual enhancement and individual differences jointly operate in bilingual subtitle contexts, this study aimed to clarify whether textual enhancement facilitates incidental vocabulary learning, and how this depends on learners' working memory, language proficiency, and vocabulary size. Results. Forty-two Chinese English majors watched a seven-minute documentary clip with bilingual subtitles in a within-subject, item-level design, in which half of the target words were visually enhanced, and half were not. Vocabulary gains were assessed through pre- and immediate post-tests measuring form recognition, meaning recall, and meaning recognition. The data were analysed using generalized linear mixed models with participants and items as random effects. Results. Textual enhancement significantly increased the likelihood of correct responses in form recognition (beta = 1.608, p < 0.001, OR = 5.00) and meaning recall (beta = 2.637, p = 0.004, OR = 13.98), but not in meaning recognition (beta = 1.074, p = 0.161). Language proficiency significantly predicted form recognition (beta = 0.506, p = 0.044), meaning recall (beta = 0.818, p = 0.014), and meaning recognition (beta = 0.646, p = 0.017). Working memory predicted form recognition (beta = 0.486, p = 0.037) and meaning recognition (beta = 0.622, p = 0.009), whereas vocabulary size predicted only meaning recognition (beta = 0.529, p = 0.030). Conclusion. The findings suggest that, in bilingual subtitle contexts, textual enhancement primarily supports form-level learning rather than deeper semantic integration. Language proficiency emerged as the most consistent predictor of incidental vocabulary gains, while working memory and vocabulary size showed more selective effects depending on the type of vocabulary knowledge assessed. These results shed light on the boundary conditions under which visual enhancement is effective in cognitively demanding multimodal input.
Background. As artificial intelligence (AI) becomes increasingly embedded in educational processes, teacher education programs must prepare future teachers to meaningfully integrate AI tools into their pedagogical practice. Despite this growing need, little is known about pre-service English teachers’ AI related pedagogical competencies, or how their attitudes toward AI align with these competencies. Addressing this gap is essential for informing curriculum design and professional development. This study contributes to the emerging fields of AI integration in language teacher education. It offers timely insights into how future English language teaching (ELT) professionals perceive and use AI, an area with limited empirical evidence. Purpose. To investigate the artificial intelligence technological pedagogical content knowledge (AI-TPACK) levels of pre-service English teachers and their attitudes toward AI and to examine whether these two constructs are meaningfully related. Method. 244 pre-service teachers in their junior or senior year from two state universities participated in this study. The participants completed two assessment scales: the AI-TPACK and the GenAI attitude scale. The obtained data were analysed via SPSS, and additional analysis was performed in the R environment. Results. The findings revealed that pre-service teachers in the ELT department reported relatively positive perceptions across AI-TPACK dimensions and demonstrated positive attitudes toward AI as measured by a learner-oriented GenAI attitude scale. Item-level mean scores across AI-TPACK dimensions ranged from 3.61 to 3.88, indicating relatively similar perceptions across the subdimensions. Notably, no significant relationship was identified between any of the constructs of AI-TPACK and attitudes. Conclusion. Although pre-service English teachers demonstrated positive attitudes towards AI as a learner and relatively positive perceptions of AI-related pedagogical competence, the absence of significant relationships between these two constructs may point to a possible attitude-competence gap in AI integration. These findings suggest that favourable perceptions of AI alone may not be sufficient for the development of integrated pedagogical competence. Therefore, teacher education programs may need to provide more structured and practice-oriented opportunities for AI integration in ELT contexts.
Background. The emergence of artificial intelligence (AI) tools such as ChatGPT has transformed English as a Foreign Language (EFL) education by offering new approaches to language learning and teaching. However, the evidence base remains limited in three important respects: prior studies often homogenize students and instructors, report inconsistent findings on key acceptance determinants, and rarely examine AI adoption through the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2) within culturally and pedagogically specific EFL contexts. Purpose. To address inconsistent and fragmented evidence on AI adoption in language education, this study examines the determinants of ChatGPT adoption among Saudi EFL students and instructors using UTAUT2, and evaluates how demographic factors condition these relationships in a generative AI context. Method. The study employed a quantitative cross-sectional survey with 345 participants (students and instructors) from four Saudi universities. Separate UTAUT2-based questionnaires were validated through pilot testing and confirmatory factor analysis. Multiple regression analyses were conducted separately for each group to examine determinant effects, with interaction terms used to test the moderating roles of gender, age, and experience. Results. Across both groups, performance expectancy was the strongest common predictor of behavioral intention (students: R = 0.42; instructors: R = 0.40). Beyond this shared effect, the patterns diverged. For students, social influence (R = 0.30) and hedonic motivation (R = 0.22) also played a significant role, whereas effort expectancy and habit did not reach significance. For instructors, by contrast, effort expectancy (R = 0.11) and habit (R = 0.10) were significant alongside performance expectancy, while social influence and hedonic motivation were not. Facilitating conditions and price value were non-significant in both groups. The moderation analysis further indicated that gender, age, and experience affected selected relationships between predictors and behavioral intention, with these effects varying across students and instructors rather than following a uniform pattern. Conclusion. Adoption of ChatGPT in Saudi EFL education depends on both shared and group-specific factors, with performance expectancy as the strongest common predictor. The results emphasize the need for culturally informed and targeted strategies that address demographic and experiential differences and provide theoretical and practical guidance for AI integration in language education.
Background. Given the rapid advances in technology, artificial intelligence (AI) offers significant potential in educational contexts by enabling personalized learning and enhancing learner engagement. However, the development of AI-related scholarship in English as a Foreign Language (EFL) education, including its publication trajectory, collaborative networks, citation patterns, and thematic distribution, remains insufficiently explored. Purpose. This study examines the extant research on AI in EFL education, spanning from the early developments in 1996 to mid-2025, with the aim of identifying structural patterns, thematic concentrations, and under-examined areas to inform future research priorities and pedagogical directions. Method: By analyzing 829 publications from the Web of Science (WoS) and Scopus databases, screened by a single researcher, and utilizing the VOSviewer software and Microsoft Excel, the study identifies and interprets significant trends, including publication trends, international collaborative networks, citation impact and thematic developments across the literature. Bibliometric indicators, country collaboration networks, journal citation patterns, and keyword co-occurrence structures were analysed; results should be interpreted in light of English-language database coverage, single-researcher screening, and threshold-dependent keyword mapping. Results. The findings reveal an exponential rise in research output since 2022, with China leading in publication volume, citation impact, and collaborative centrality. Highly impactful journals demonstrate substantial engagement across disciplinary areas, with Computer Assisted Language Learning emerging as the most influential source. The co-occurrence analysis of Keywords highlights a concentration on generative AI tools, particularly ChatGPT, and AI-enhanced writing, while areas such as speaking, listening, reading, and self-regulated learning appeared underrepresented in the keyword structure of the corpus, suggesting potential gaps in the thematic scope of current AI-related EFL research. Conclusion. This study provides the first bibliometric account of AI in EFL education to encompass the generative AI period, simultaneously mapping collaboration networks, citation trajectories, and keyword structures. The findings offer evidence-based insights into research trends and thematic imbalances, with implications for future inquiry and pedagogical innovation in AI-enhanced EFL education.
Background. While videos increasingly support flexible and student-centred learning in higher education, their pedagogical potential is often limited in practice. Transient information effects and insufficient emphasis on generative processing constrain active cognitive engagement. Empirical evidence evaluating how specific pedagogical frameworks, such as the 5E guided inquiry model, impact cognitive load dimensions in EFL higher education remains limited. Purpose. To investigate how guided inquiry can foster generative processing in video-based learning and help address the transient information effect in terms of learning outcomes and cognitive load. Method. A pre/post-test quasi-experiment was conducted with 135 English department students, divided into a control group (n=65) and a treatment group (n=70). The participants were first-year EFL students who were novices in the subject domain and reported varied English proficiency, most commonly intermediate or upper-intermediate. Both groups studied the same content across eight sessions. The control group used educational videos and direct instruction, while the treatment group used educational videos combined with the 5E model of guided inquiry. Cognitive load was measured across essential, extraneous, and generative processing, while pre- and post-tests assessed retention and transfer. Results. Findings demonstrate that the treatment group achieved significantly higher retention (p < .001, r = .414). Due to a ceiling effect, transfer outcomes were deemed uninterpretable rather than evidence of comparability. While the aggregate cognitive-load subscales did not show between-group significance, results demonstrated a localised effect on invested cognitive effort, with one item remaining significant after Holm-Bonferroni correction (p < .001, r = .326). Within this context, the findings suggest that the initial increase in extraneous processing was not sustained over time and appeared to diminish as students adapted to the guided inquiry structure. Conclusion. This study applies multi-dimensional cognitive load measurements to a 5E guided inquiry intervention within an EFL higher education context. The results indicate that the original hypothesis was supported for retention and partially supported, at the item level, for invested cognitive effort. These findings offer a context-bound perspective for curriculum designers, suggesting that structured inquiry prompts can support foundational knowledge consolidation and specific facets of engagement in comparable EFL multimedia learning environments.
Background. Usage examples are an important feature of pedagogical dictionaries, as they help learners understand meaning, context, and lexical use. Although web-based resources and large language models offer new possibilities for generating dictionary examples, their pedagogical suitability has not yet been sufficiently examined in Modern Greek. Purpose. To address this gap, this exploratory study compares Web-based and AI-generated usage examples across three lexically challenging categories in Modern Greek: polysemous verbs and nouns, and idiomatic expressions. It examines whether and how the pedagogical suitability of each source varies according to lexical complexity for learners aged 11-15 and explores their potential integration in pedagogical lexicography. Materials and Methods. A small-scale exploratory dataset of 81 usage examples was collected from Web sources and generated by GPT-5.3 Instant and DeepSeek-V3 and was evaluated using a combined framework drawing on GDEX criteria and pedagogical appropriateness indicators on a three-point ordinal scale. A learner validation phase with 41 students (aged 11-15) was also conducted. Results. Within the analysed dataset, the findings suggest category-related differences across lexical categories. Web-derived examples showed greater semantic diversity and higher correct meaning identification rates in the learner evaluation (56.50%), particularly for figurative and idiomatic uses. AI-generated examples, especially those produced by GPT-5.3, showed higher intelligibility scores and performed more consistently for semantically stable items such as polysemous nouns, but exhibited limitations in idiomatic language, with correct meaning identification for GPT-5.3 dropping to 32.37% for specific idiomatic items. DeepSeek-generated examples showed the most pronounced limitations within the analysed categories, particularly in typicality and pedagogical appropriateness. Conclusion. The findings suggest that, within the analysed dataset, the pedagogical value of usage examples depends not only on their source, but also on the type of lexical complexity involved. Neither Web-based nor AI-generated examples can independently meet the requirements of a learner-oriented dictionary. Instead, the findings support a hybrid category-sensitive approach that combines Web-based authenticity, expert curation, selective AI assistance, and attention to learners’ actual comprehension. The methodological framework developed here may offer a basis for future comparative work in other morphologically complex and idiomatically rich languages, though transferability remains to be tested.
Background. Research on EFL writing has widely examined formative feedback and writing self-efficacy, but fewer studies have integrated feedback quality, task-specific writing self-efficacy, and perceived writing competence within a single empirical model. The unresolved problem is whether students' confidence in performing academic writing tasks functions as an intermediary mechanism through which feedback perceptions are associated with competence self-evaluations, particularly in underrepresented higher education contexts where formative assessment practices are still developing. Purpose. This study clarifies whether writing self-efficacy operates as a linking construct between formative feedback and perceived writing competence among EFL university students in Kosovo. In doing so, it aims to extend social-cognitive explanations of academic writing by showing how an instructional condition may be connected to students'self-evaluativejudgments through writing-related capability beliefs. Method. Self-report data were collected from 400 students enrolled in English-medium academic writing courses at four higher education institutions in Kosovo. The measurement model was examined through confirmatory factor analysis, and the hypothesized relationships were tested through observed-variable path analysis. Because the study was cross-sectional and based on self-report data, all findings are interpreted as associative rather than causal. Results. The hypothesized pattern was supported. Formative feedback was positively associated with writing self-efficacy (R = 0.39) and perceived writing competence (R = 0.30), while writing self-efficacy was positively associated with perceived writing competence (R = 0.37). The indirect association through writing self-efficacy was significant (R = 0.15), indicating partial statistical mediation. Substantively, students who perceived feedback as clearer and more usable also tended to report stronger writing confidence, which in turn was linked to more positive competence evaluations. Conclusion. The study contributes bytesting an integrated model of feedback, self-efficacy, and perceived writing competence in an underrepresented Balkan EFL higher education context. It positions writing self-efficacy as an intermediary construct that helps explain why formative feedback is associated with students' competence self-evaluations. The findings support the broader applicability of Social Cognitive Theory to EFL writing while also highlighting the need for cross-cultural research that compares how feedback traditions and student engagement shape this model across educational systems.
Introduction. Epistemological transparency remains a persistent problem in educational research: studies often report methods in detail but leave unstated the assumptions about knowledge, evidence, and validity that justify those methods. This article examines how epistemological stances are articulated in high-impact research on higher education and how they correspond to methodological choices in original research articles and literature reviews. Method. The study is based on qualitative content analysis of 100 Scopus-indexed publications on higher education published in 2021-2025: 50 original research articles and 50 literature reviews, each cited at least 50 times. The coding framework was developed iteratively and covered lexical markers, methodological signals, quality indicators, and citational practices. These indicators were used to identify four epistemological orientations: post-positivist, interpretivist/constructivist, pragmatist, and critical/transformative. Inter-coder reliability was assessed using Cohen's kappa and Krippendorff's alpha. Results. Explicitepistemological positioning wasfound in onlyone publication.Acrossthe corpus, post-positivist orientations were most frequent (53%), followed by interpretivist/constructivist (33%), pragmatist (13%), and critical/transformative orientations (1%). The distribution differed sharply by publication type. Reviews were predominantly post-positivist (74%), with smaller shares of interpretivist/constructivist (24%) and pragmatist (2%) orientations. Articles showed a more balanced profile: interpretivist/constructivist orientations accounted for 42%, post-positivist for 32%, pragmatist for 24%, and critical/transformative for 2%. Reviews signalled stance mainly through systematic procedures, including PRISMA-based protocols, search strategies, quality appraisal, and reliability checks. Articles more often signalled stance through theoretical framing, methodological design, and the treatment of participants' perspectives or practical outcomes. Borderline cases showed that methodological indicators alone cannot always establish epistemological stance, especially where systematic procedures are combined with interpretive, pragmatist, or critical aims. & Scy;onclusion. Epistemological stance is usually present in high-impact educational research, but it is rarely named. This implicitness leaves readers to reconstruct the philosophical basis of a study from methodological and rhetorical cues. Greater transparency does not require authors to declare allegiance to a single paradigm. It requires them to explain how their assumptions about knowledge and evidence shape the formulation of research questions, the selection of methods, the interpretation of findings, and the limits of the claims made.
Background. The study highlights a gap in research on politeness in EFL academic interaction, noting that previous studies focused on classroom discourse analyses. It emphasizes the need to examine real interaction during thesis defenses, particularly the influence of institutional power on sequential politeness practices in high-stakes assessment. Purpose. This study examines the distribution and negotiation of politeness strategies between examiners and candidates during EFL thesis defenses, highlighting the interactional construction of institutional authority and face management in academic assessment. Method. The study combines quantitative and qualitative analysis. It is based on eight undergraduate thesis defense sessions involving eleven examiners and eight candidates. The analysis covers 772 utterances coded according to Brown and Levinson's politeness framework. The quantitative results show how politeness strategies were distributed across participant roles, while the sequential discourse analysis shows how these strategies were used and negotiated during the defense interaction. Since the utterances come from the same speakers and sessions, the statistical findings are interpreted cautiously and used mainly to identify patterns in the data. Results. The findings revealed patterned variation in politeness strategy use between examiners and candidates, chi(2)(3) = 23.94, p < .001, with a moderate association (Cram & eacute;r's V = .346). Data showed that negative politeness was the most dominant strategy (58.5%). Candidates relied heavily on negative politeness (73%) to minimize imposition and display deference, whereas examiners employed a more varied distribution of strategies, including positive politeness (34%) and bald-on-record forms (20%), to manage evaluation, procedural control, and institutional authority. Sequential analysis further showed that examiner authority was most visible during questioning and challenge phases through turn management, epistemic stance-taking, interruptions, and repair sequences. Conclusion. In summary, these results demonstrate that politeness strategies function as interactional resources for negotiating institutional power and evaluative relationships in EFL thesis defenses. Thus, the findings contribute empirically to research on institutional discourse and pedagogically support the inclusion of pragmatic preparation in EFL thesis defense training.
Background: In spite of intense interest, no quantitative, model-based comparison of semantic distances in verb aspectual triplets has been conducted at scale. Also, no statistical estimate exists for how prefixes contribute (or fail to contribute) to semantic identity. Purpose: To obtain objective quantitative evaluation of the semantic distance between the verbs in triplets: the basic imperfective verb, the perfective verb, and the secondary imperfective verb. The main problem we solve is which of the two imperfective verbs is semantically closer to the perfective verb, i.e. which pair is aspectual. Method: The present study covers 439 verb aspectual triplets from the database by L. Janda. We apply the computational technique of embeddings (vector semantics) to solve the issues. Two complementary embedding models are employed: a static word embedding model (fastText) trained on the Russian subcorpus of the Google Books Ngram database and a contextual sentence embedding model (LaBSE). Results: Both embedding models consistently show perfective verbs to be semantically closer to secondary imperfective verbs than to basic imperfective verbs. The effect is moderate for fastText (Cliff>s Delta = 0.33, 95% bootstrap CI [0.26, 0.40]) and large for LaBSE (Cliff>s Delta = 0.62, 95% bootstrap CI [0.56, 0.68]). The Wilcoxon signed-rank test confirmed high statistical significance for both models (fastText: W = 9,691, p < 10-47; LaBSE: W = 23,464, p < 10-20). 12 of 15 prefixes (the LaBSE model) and 8 of 15 (the fastText model) showed significant effects after False Discovery Rate correction. All effect size estimates are reported with 95% bootstrap confidence intervals based on 5,000 resampling iterations. Conclusion: Our findings confirm that Russian verbal prefixes carry non-empty semantics, with systematic semantic differences between aspectual and purely derivational pairs. The nonoverlapping confidence intervals between models indicate that LaBSE captures the semantic distinction between aspectual and semantic verb pairs more robustly than the static fastText embeddings. The study establishes a quantitative framework for analyzing derivational semantics and reveals significant heterogeneity in prefix semantic contributions.
Introduction: Rapid advances in artificial intelligence are reshaping higher education and intensifying debate about the effectiveness, risks, and ethical implications of AI supported learning and academic writing. Yet faculty experiences with AI for research writing remain comparatively underexamined. Purpose: To examine university faculty members’ experiences and perceptions of using AI tools to support research writing and publication practices in higher education, with particular attention to issues of AI literacy and responsible use.. Method: This exploratory design based case study reports two one day immersion courses delivered through the university writing center at HSE University in Fall 2023 and Spring 2024. Participants were two cohorts of trainees (Case 1 n = 19; Case 2 n = 15) and two course instructors. Data comprised post course trainee feedback covering course usefulness, content, assignments, activities, instructor performance and feedback, and self reported involvement, as well as semi structured interviews with instructors about teaching and writing with AI tools. Numeric feedback items were analyzed using descriptive statistics in R. Interview data were analyzed using thematic analysis focused on key takeaways, concerns, and suggestions for improvement. Results: Trainees rated the courses highly, especially the practice oriented format (Case 1 mean = 9.4 out of 10; Case 2 mean = 9.5). Instructors corroborated strong engagement but reported that an eight hour single day schedule increased cognitive load and reduced pedagogical stamina. Ethical concerns differed across cohorts: responsible AI use was discussed extensively in Case 1, whereas Case 2 participants mainly acknowledged the need for ethical use. Participants also demonstrated heterogeneous AI literacy and limited prior exposure to AI tools, reported by seven trainees in Case 1 and six trainees in Case 2. Conclusion: Findings informed immediate redesign of the writing center’s offerings by strengthening attention to AI ethics and literacy, extending course formats, and maintaining hands on learning as a core principle. Course designers should consider participants’ prior experience, ensure reliable access to a curated set of tools, allocate time for guided familiarization and task completion, and plan for classroom management in mixed literacy groups. The study contributes faculty focused evidence to discussions of AI mediated research writing and motivates experimental testing of AI integrated pedagogical frameworks, including Intelligent TPACK.
Introduction: Oral corrective feedback (OCF) is widely recognized as crucial for second language learning, yet its effectiveness depends significantly on learners' receptivity and beliefs. Existing instruments for measuring OCF beliefs face methodological limitations: they rely on abstract, theory-driven terminology that may be misinterpreted by learners, impose researcher-generated factor structures that may not reflect authentic learner perspectives, and carry culture-bound interpretive burdens that limit cross-cultural comparability. Purpose: This study develops and validates a scenario-based instrument for measuring learner beliefs about OCF that prioritizes accessibility and ecological validity by presenting concrete classroom situations rather than abstract terminology. The two-phase validation approach employs exploratory factor analysis (EFA) to identify how learners naturally organize their beliefs, followed by confirmatory factor analysis (CFA) to validate the discovered structure. Method: A 50-item scenario-based questionnaire was developed through expert review and pilot testing, then administered to 668 B2-level English learners in a Turkish university preparatory program. The sample was randomly split: EFA was conducted on the first subsample (n = 334) using principal axis factoring with oblimin rotation, and CFA was performed on the holdout sample (n = 334) to validate the factor structure. Results: EFA identified a six-factor structure explaining 61.4% of variance: Affective Response to Correction (strongest dimension, α = .85), OCF Type Preferences (α = .87), Correction Timing Preferences (α = .81), Context-Sensitive Correction (α = .79), Correction Source Preferences (α = .72), and OCF Uptake and Response (α = .76). CFA confirmed the six-factor solution with acceptable fit indices (CFI = .91, TLI = .90, RMSEA = .06, SRMR = .07). Notably, learners organized correction types by experiential familiarity rather than theoretical distinctions (input-providing vs. output-prompting), and affective considerations emerged as central rather than peripheral to belief systems. Conclusion: The scenario-based approach yields accessible items that reduce culture-bound interpretive burden and enhance cross-cultural comparability. The identified dimensions—affect, type, timing, context, source, and uptake—map onto decision points common to EFL/ESL classrooms globally, offering a scalable measurement model for comparative research across educational systems. Findings encourage pedagogy that is adaptable to context, sensitive to affect, and aligned with how learners actually experience correction. Future research should test measurement invariance across diverse cultural and linguistic settings to establish international applicability.
Background: Although foreign language learning boredom (FLLB) is known to negatively influence learners' L2 (second/foreign language) engagement and willingness to communicate (L2WTC), prior studies have reported inconsistent effect sizes and limited theoretical integration. Purpose: To (1) quantitatively estimate the overall strength of the associations between FLLB and both L2 engagement and L2WTC, and (2) identify significant moderators that might account for variability across studies. Method: A correlational meta-analysis was conducted, including 14 studies on FLLB-L2 engagement and 11 studies on FLLB-L2WTC. Effect sizes were computed using random-effects models and transformed via Fisher's Z, and moderator analyses of meta-regression and subgroup analyses were performed to examine the potential moderator impacts of learner demographics, research context, and measurement tools. Results: The results indicated a medium-to-large negative correlation between FLLB and L2 engagement (r = -0.506) and a small-to-medium negative correlation with L2WTC (r = -0.302). Moderator analyses revealed that research context and engagement measurement scale significantly moderated the FLLB-L2 engagement relationship, while gender ratio and L2WTC measurement scale significantly moderated the FLLB-L2WTC relationship. Conclusion: These findings underscore the need for reducing boredom to enhance learner engagement and communicative willingness in L2 contexts. Future research should further explore individual and contextual moderators, particularly among younger learners and in digital learning environments.
Background: The rapid global expansion of English-Medium Instruction (EMI) has transformed higher education, particularly in business programs that require students to master complex disciplinary knowledge while overcoming language-related barriers. EMI students must simultaneouslyovercome linguistic, cognitive, and cultural barriers, which can increase academic stress, reduce participation, and negatively impact learning outcomes. Despite these challenges, most EMI research has relied on correlational models emphasizing self-efficacy, motivation, or language proficiency, while overlooking psychological constructs that capture students' abilityto manage daily academic stressors. Academic buoyancy-the capacity to persist through routine academic setbacks-offers a theoretically robust and practically relevant lens for understanding student adaptation in multilingual learning environments. Purpose: This study examines academic buoyancy as a mediating mechanism linking teaching effectiveness to deep learning outcomes in EMI business education. Drawing on academic resilience theory, interaction theory, and self-regulated learning theory, we investigated predictors of academic buoyancy and tested whether it mediates the relationship between teaching effectiveness and deep learning. Method: Data were collected via cross-sectional surveyfrom 215 international students enrolled in EMI business courses at a South Korean university. Multiple regression analysis identified predictors of academic buoyancy, and Hayes' PROCESS macro with 5,000 bootstrap samples tested mediation pathways. Results: Learner-professor interaction (beta = .163, p < .001), teaching effectiveness (beta = .146, p < .001), and cognitive engagement (beta = .102, p = .006) significantly predicted academic buoyancy, collectively explaining 30.6% of the variance. Mediation analysis demonstrated that academic buoyancy fully mediated the relationship between teaching effectiveness and deep learning (indirect effect = .071, p < .05; 95% CI [.012, .130]), with no significant direct effect (beta =-.023, p > .05). Notably, peer interactions, content engagement, and metacognitive self-regulation did not significantly predict academic buoyancy after controlling for other variables. Conclusion: Academic buoyancy served as a key psychological mechanism through which teaching effectiveness influences deep learning in EMI contexts. The findings suggest that effective EMI pedagogy operates primarily by fostering student resilience rather than through direct content transmission, with practical implications for faculty development emphasizing both relational and pedagogical dimensions of instruction.
Background: Although oral corrective feedback (OCF) has been shown to benefit second language acquisition, relatively few studies have examined students' attitudes toward it, which may impact on its effectiveness. Moreover, even though students' attitudes to language learning can depend on different variables, such as students' level or cultural background, very few studies have looked at the possible impact of such variables on students' attitudes to OCF. Even fewer studies have focused on affective responses to OCF and hardly any have included non-WEIRD populations. Purpose: To compare EFL students' attitudes toward immediate oral corrective feedback and their affective responses to it across two contrasting higher-education contexts (Spain and Algeria) using a single standardized questionnaire. Method: The present study compared students' attitudes to OCF, as well as their affective responses to OCF, in two contrasting EFL higher-education settings with different L2 teaching traditions, by means of a cross-sectional Likert-type questionnaire: 213 Spanish and 261 Algerian EFL students at the university level. Descriptive statistics in the form of percentages of agreement and disagreement were calculated, and chi-square tests of independence were carried out to investigate possible differences in attitudes. Results: Contrary to previous findings that cultural background only appears to have a minor effect on students' attitudes to OCF, we found significant differences between Spanish and Algerian students' attitudes for several aspects of OCF, such as frequency, error type, OCF-type and affective responses to OCF. For instance, although most students expressed a wish to receive as much OCF as possible, there were more Algerian students who preferred to be corrected only in certain cases. With regard to OCF types, the majority of students in both contexts expressed a preference for metalinguistic feedback, but Algerian students were more positive about recasts. Conclusion: Given these observed differences, we cannot expect students from all contexts to have similar beliefs about OCF. Teachers are thus advised to discuss OCF with their students in order to avoid mismatches between students' beliefs and teachers' practices, as this can negatively affect the learning process. This study sheds light on the beliefs about OCF of students from different backgrounds, including an underrepresented, non-Western population, i.e., Algerian EFL students. It also highlights a need to further explore the possible reasons behind the observed differences in attitudes.
Background: Although both peer and AI-generated feedback are increasingly used in EFL writing instruction, little is known about how they differ in terms of cognitive depth and interpersonal delivery. Existing studies often overlook the mechanisms through which feedback operates, limiting instructors' ability to design balanced feedback ecosystems. Purpose: To provide a comparative analysis of peer and ChatGPT-4o feedback on undergraduate EFL academic writing, drawing on Hattie and Timperley's (2007) cognitive model and Hyland and Hyland's (2006) interpersonal feedback strategies. It aims to identify how each source targets task, process, and self-regulation levels, and how their rhetorical styles shape learner engagement. Method: Thirty Iranian undergraduate EFL students participated in a qualitative classroom-based study in which each essay received both peer and ChatGPT-4o feedback. Atotal of 430 peer comments and 224 ChatGPT feedback units were coded deductively using validated analytical frameworks. Inter-rater reliabilityfor 20% of the dataset yielded substantial agreement (kappa = .82). Results: Peer feedback was subjective, socially expressive, and frequently mitigated, with comments focusing on sentence-level issues but also including self-regulation prompts (15%) that encouraged reflection and decision-making. ChatGPT-4o provided predominantly task-level feedback (94%), characterized by structured, consistent, and objective guidance on grammar, organization, and coherence. However, its feedback exhibited minimal interpersonal variation and rarely promoted metacognitive engagement. Conclusion: The findings indicate that peer and AI-generated feedback serve complementary pedagogical functions: AI offers technical accuracy and consistency, while peer feedback contributes emotional support and occasional reflective prompts. A hybrid feedback model that integrates both sources may therefore enhance the revision process in EFL writing instruction.
Background: Metaphors are found to have positive impacts on vocabulary teaching and learning. However, further investigations into the existing literature disclosed a lack of thorough research regarding the utilization of conceptual metaphors in teaching business vocabulary. In addition, while these studies have shed light on the impact of orientational metaphors in teaching vocabulary, there remains a lack of consensus concerning the effects of teaching vocabulary through structural metaphors or ontological metaphors. Purpose: To examine the effectiveness of utilizing the Conceptual Metaphor Theory (CMT) in teaching English vocabulary for business purposes to enhance students' vocabulary learning outcomes, including vocabulary acquisition, retention, and lexical inference. Method: A quasi-experimental study was designed and conducted over a duration of 10 weeks with the participation of 58 second-year students specializing in International Business. The participants were randomly and equitably allocated to the experimental group, which utilized the CMT-based teaching method, and the Control group, which employed traditional teaching methods. The data were collected via a pre-test, immediate post-test, and delayed post-test, thereafter, analyzed using IBM SPSS software version 26.0. Results: While the pre-test scores exhibited their homogeneity in their English proficiency before the experiment, the students in the experimental group achieve a mean score of 7.662; 1.228; and 7.429, surpassing the control group's mean score of 6.710; 0.510; and 6.476 regarding the results gained in the overall Immediate post-test, Inference questions, and Delayed post-test respectively. Conclusion: CMT-based teaching approach significantly enhances students’ vocabulary learning outcomes, including vocabulary acquisition, lexical inference, and vocabulary retention. The current study contributes to the literature relating to vocabulary instructional approach. The study indicates that instructors can facilitate business vocabulary learning by explicitly emphasizing the conceptual frameworks that support key lexical elements. Furthermore, the findings suggest that future classroom applications may be improved by integrating metaphor education into case studies, presentations, and real-world business situations, enabling students to utilize metaphors in contextually relevant manners. These pedagogical insights can assist educators in designing more cognitively engaging evidence-based practice programs that promote long-term vocabulary growth.
Background: For a long time, concepts such as the intercultural communicative competence have challenged the view of culture as purely fact-based, objective content to be transmitted to students, from whom only retention is expected. Today, cultural knowledge plays virtually no role in international standardised English exams such as TOEFL or the Cambridge English Qualifications. It is unknown to which extent this is true for examinations at the national level, and the materials used to prepare students for these examinations. Methods: The study provides a quantitative account of exam topics in the speaking part of the final exam in Czech grammar schools. The sample consists of 206 grammar schools in the Czech Republic (58% of all such schools). The percentage of knowledge-oriented, culture-specific topics was calculated for each school. The study further analyses the content of four books used to prepare students for the exam. The sample comprises four frequently used preparatory books. The text in the chapters that deal with cultural topics is classified using Larzén’s (2005) typology of culture. Results: The study has demonstrated that approximately 50% of the final exam topics are focused on specific English-speaking countries, their geography, history and literature. A similar tendency, but significantly more pronounced, has been found in the preparatory materials. More than 95% of their content was heavily knowledge-oriented with no discernible difference between older and recently published materials. Virtually no evidence of a skill-based or intercultural approach to culture was found. Conclusions: The study has confirmed what is often only an unfounded assumption by researchers and it has shown that the traditional factual view of culture is still immensely popular with no change in sight. Recommendations for material writers, teachers, and teacher educators are provided at the end of the study.
