WHEN STUDENTS WITH DISABILITIES STRUGGLE TO SPEAK ENGLISH: RECLAIMING EMPATHY THROUGH AI-ENHANCED INCLUSIVE SCHOOL CULTURE 

Abstract ID: 167 

Ramli 

Universitas Islam Internasional Indonesia 

Corresponding email: ramli@students.uiii.ac.id 

This qualitative case study investigates how empathy-driven school culture and inclusive policy shape English language learning for students with disabilities, with a specific focus on speech delay, in an Indonesian inclusive primary school. Situated at SD Semut-Semut in Depok, the study employs semi-structured interviews, classroom and school-wide observations, and document analysis to examine (a) teachers’ and principals’ understandings of empathy, (b) practices through which empathy is institutionalized in school culture, and (c) the implications of empathy-informed culture for advocacy-oriented policy and classroom support for English speech development. Findings indicate that empathy is conceptualized as a multidimensional professional competence cognitive, affective, and compassionate rather than as pity or charity. This competence is embedded in individualized education programs, flexible assessment practices, and collaborative arrangements between classroom teachers and GPK (shadow teachers), which together scaffold participation for learners with speech and language delays. Empathy-driven practices manifest in simplified instructions, differentiated indicators of achievement, and emotionally secure classroom climates that encourage oral engagement in English. At an institutional level, empathy is sustained through structured coordination, continuous inclusive training, and quality control mechanisms that align the school’s philosophy of “everybody is unique” with everyday pedagogical decision-making. The study argues that institutionalizing empathy as an ethical and cultural framework is critical for bridging the persistent gap between inclusive policy and practice and for advancing targeted, sustainable support for English speech development among students with disabilities. 

Keywords: Empathy; Inclusive; Speech Delay; disabilities; school culture.

STUDENTS’ PERCEPTIONS OF AI CHATBOTS AS SPEAKING PARTNERS AND THEIR SOCIAL COMPETENCE IN EFL SPEAKING PRACTICE 

Abstract ID: 156 

Noor Aida Aflahah, Hafizhatu Nadia, Elsa Marina Desiarti 

Universitas Muhammadiyah Banjarmasin 

Corresponding email: nooraida_aflahah@umbjm.ac.id 

Artificial Intelligence (AI) chatbots have become increasingly popular in English as a Foreign Language (EFL) learning by providing interactive and flexible opportunities for speaking practice. While previous studies have demonstrated their effectiveness in improving linguistic performance, limited research has explored their contribution to students’ social competence. This study aimed to investigate EFL students’ perceptions of AI chatbots as speaking partners and examine their perceived social competence after using AI chatbots for English speaking practice. A descriptive quantitative design was employed involving 74 undergraduate EFL students. Data were collected using a validated and reliable questionnaire comprising 28 items. The findings showed that ChatGPT (75.7%) and Gemini (62.2%) were the most frequently used AI chatbots for speaking practice. Overall, students expressed positive perceptions toward AI chatbots. More than 80% agreed that AI chatbots encouraged more frequent speaking practice, improved speaking fluency, provided meaningful conversational opportunities, and enabled learning beyond the classroom. Students also perceived AI chatbots as easy to use, responsive, and capable of providing contextually relevant interactions. Regarding social competence, most participants reported confidence in initiating and maintaining conversations, understanding conversational intentions, following conversation flow, and adapting their communication to different contexts. However, lower levels of agreement were found in their ability to respond appropriately, express ideas clearly, and organize their thoughts in English conversations. These findings indicate that AI chatbots can serve as effective speaking partners while supporting the development of students’ social competence. Nevertheless, pedagogical guidance remains essential to enhance higher-order communicative skills in AI-assisted EFL learning. 

Keywords: Artificial Intelligence, AI Chatbot, EFL, Speaking Practice, Social Competence

SAME TEACHERS, NEW TECHNOLOGY: A TWO-WAVE LONGITUDINAL CASE STUDY OF INDONESIAN EFL TEACHERS’ SELF-EFFICACY IN THE GENERATIVE AI ERA 

Abstract ID: 243 

Aulia Rahmawati 

Universitas Muhammadiyah Palembang 

Corresponding email: auliarahmawati.edu@gmail.com 

Teacher self-efficacy in technology integration is widely recognized as a key factor influencing teachers’ responses to technological change. However, little is known about how the same teachers’ self-efficacy develops across successive technological transitions. This study investigated how Indonesian EFL teachers’ technology integration self-efficacy, first documented during the COVID-19 emergency remote teaching period in 2022, evolved as those same teachers began integrating generative artificial intelligence (Gen AI) into their teaching in 2026. A two-wave longitudinal qualitative case study was conducted with five of the six teachers who participated in the original study. Data were collected using the Generative AI Self-Efficacy (GAISE) questionnaire and semi-structured interviews. The questionnaire showed that all participants demonstrated high levels of pedagogical and critical/evaluative confidence. However, the interview findings revealed that confidence in operating AI tools was derived from prior technology integration experience, whereas confidence in integrating AI into teaching and in critically evaluating AI-generated content developed through new learning and professional experience, regardless of teachers’ initial technology integration self-efficacy. Participants also reported limited AI-specific professional development, the absence of institutional AI policies, and concerns about academic integrity as continuing challenges. These findings suggest that prior technology integration self-efficacy provides an important foundation for adopting Gen AI but is insufficient on its own to support effective pedagogical and evaluative use of AI. The study contributes longitudinal evidence on the development of teacher Gen AI self-efficacy. It highlights the importance of professional development that strengthens teachers’ pedagogical and evaluative capacities alongside their technical AI skills. 

Keywords: AI-TPACK; EFL teachers, teachers’ self-efficacy; generative AI; two-wave longitudinal study 

HUMAN-CENTERED OPEN AND DISTANCE LEARNING: SUSTAINING STUDENTS THROUGH THE EXPERIENCE OF UNIVERSITAS TERBUKA, INDONESIA 

Abstract ID: 162

Widya Rizky Pratiwi 

Universitas Terbuka 

Corresponding email: widya_pratiwi@ecampus.ut.ac.id 

The rapid expansion of digital technologies has transformed open and distance learning (ODL), increasing access to higher education across diverse geographical and socio-economic contexts. However, access alone does not ensure student resilience, engagement, or well-being. This paper explores how a human-centered approach to ODL sustains students through the experience of Universitas Terbuka, Indonesia, one of the world’s largest open universities, serving more than 767,000 students across 39 regional learning centers and one overseas learning service. Using an institutional case study and reflective practice approach, the paper examines how student resilience is fostered through flexible learning pathways, responsive tutoring, multi-channel academic support, AI-assisted services, and continuous student feedback. Rather than positioning technology as the primary driver of learning success, the analysis highlights the importance of designing learning ecosystems that accommodate students’ diverse learning conditions, professional responsibilities, geographical challenges, and well-being. The findings indicate that student resilience is shaped not only by learners’ adaptability but also by institutional systems intentionally designed to be flexible, responsive, and human-centered. Institutional evaluations further show that while approximately 87% of students report satisfaction with university services, they consistently value timely communication, meaningful feedback, responsive academic support, and accessible information as key contributors to persistence and well-being. This paper argues that sustaining students in ODL requires a shift from technology-centered implementation to human-centered system design, where digital technologies strengthen rather than replace meaningful human interaction. The Universitas Terbuka experience offers practical insights for developing more inclusive, resilient, and sustainable learning ecosystems across diverse higher education contexts.

Keywords: human-centered learning, , open and distance learning, , student resilience,, student well-being, , digital education

THE IMPLEMENTATION OF FUTURE LEARNING ECOSYSTEMS IN THE DIGITAL ERA USING HUMAN-CENTERED DIGITAL TEACHING PRACTICES (HCDTP) MODEL 

Abstract ID: 143

Ratu Sarah Pujasari, Asri Siti Fatimah, Santiana

Universitas Siliwangi

Corresponding email: ratusarah@unsil.ac.id

This study introduces and examines Human Centered Digital Teaching Practices (HCDTP), a learning ecosystem model that integrates digital technology with humanistic approaches to sustain student centred education in the digital era. The research aims to advance institutional innovation in human development and competitiveness, particularly within the strategic framework of Universitas Siliwangi. Applying a description of qualitative single case study design, this research conducts an in-depth exploration of how HCDTP principles are enacted within a specific educational context in Technology Enhanced Language Learning Course for the fifth grade at Universitas Siliwangi, Tasikmalaya, West Java, Indonesia. The HCDTP model is conceptualised around five core principles (1) human centred values, emphasizing empathy toward students’ emotional and educational needs; (2) digital learning ecosystem, promoting iterative design and real world relevance; (3) pedagogical practices, fostering collaborative and shared learning experiences; (4) human technology interaction, ensuring inclusive and accessible education; and (5) learning outcome, enhancing educational effectiveness through adaptive digital solutions. This research contributes to the growing discourse on socially responsible educational technology and offers a practically grounded model for future digital learning ecosystems.

Keywords: digital learning ecosystem, human-centered education, human-centered digital teaching practices, qualitative case study, technology integration

UNDERSTANDING UNDERGRADUATE STUDENTS’ ADOPTION AND RESPONSIBLE USE OF ARTIFICIAL INTELLIGENCE IN THESIS WRITING: INTEGRATING THE THEORY OF PLANNED BEHAVIOR AND PROBLEM BEHAVIOR THEORY 

Abstract ID: 147 

Dodi Zain

Universitas Muhammadiyah Purwokerto 

Corresponding email: zaindodi777@yahoo.com

The increasing use of artificial intelligence (AI) has transformed undergraduate thesis writing, creating opportunities while raising concerns about responsible academic use. Guided by the Theory of Planned Behavior and Problem Behavior Theory, this study explores the factors influencing students’ intention to use AI, how AI is integrated throughout the thesis-writing process, and how students balance its benefits with ethical and cognitive concerns. A qualitative phenomenological approach was employed with 12 undergraduate English as a Foreign Language (EFL) students. Data were collected through semi-structured interviews and analyzed using thematic analysis. The findings indicate that students’ intention to use AI was driven by positive perceptions of its usefulness, confidence in using AI tools, and support from their academic environment. Participants strategically integrated AI into multiple stages of thesis writing, including idea generation, literature searching, outlining, drafting, language refinement, grammar checking, and paraphrasing. Although AI enhanced writing efficiency, linguistic quality, confidence, and cognitive support, students also recognized risks related to inaccurate information, overreliance, plagiarism, and limited functionality of free AI tools. To mitigate these risks, participants adopted protective strategies, such as verifying AI-generated information with scholarly sources, comparing outputs across multiple AI platforms, and exercising critical judgment before incorporating AI-generated content into their work. These findings suggest that AI functions as a valuable but fallible academic partner, underscoring the need to integrate AI literacy and responsible AI practices into higher education to promote effective and ethical thesis writing.

Keywords: artificial intelligence, thesis writing, undergraduate students, Theory of Planned Behavior, Problem Behavior Theory, AI literacy

INTEGRATING THE 5D THINKING ESSENTIALS INTO ENGLISH TEACHER EDUCATION: A CONTENT ANALYSIS OF PRAGMATIC FEATURES IN LEARNING MODULES 

Abstract ID: 149 

Hidayah Nor , Adinda afifah anwar, M.Sayid Wijaya

UIN Antasari Banjarmasin

Corresponding email: hidayahnor@uin-antasari.ac.id 

Pre-service English teachers need to have instructional materials that foster higher-order thinking and communicative competence in order to prepare them for today’s language classrooms. Although there is growing attention to transformative teacher education, there is little understanding of the integration of critical thinking with pragmatic aspects of language use in English learning modules. The study examines the integration of the 5D Thinking Essentials into English learning modules for pre-service English teachers and the encouragement of pragmatic competence in the modules. This study reports on the English learning module developed for pre-service English teachers using qualitative content analysis. The analysis is concerned with the integration of the five dimensions of thinking, as well as pragmatic features like speech acts, politeness strategies, contextual language use, intercultural communication, and authentic communicative tasks. The data are systematically coded and categorized, and interpreted thematically to identify recurring pedagogical patterns and instructional emphases. The analysis shows that the modules include a number of elements of the 5D Thinking Essentials, such as reflective activities, inquiry-based activities, collaborative learning, and problem-solving activities. These results provide opportunities for enhancing the convergence of transformative thinking with pragmatic language instruction. This study contributes to English teacher education through the proposal of a framework of instructional material design that can simultaneously develop higher-order thinking and pragmatic competence, thereby better preparing the pre-service English teachers to be able to teach language effectively and contextually. 

Keywords: 5D Thinking Essentials, English Teacher Education, Learning Modules

HUMAN COMPETENCE IN THE AGE OF MACHINE TRANSLATION: CHALLENGES FOR TRANSLATION LEARNERS 

Abstract ID: 145

Pebri Prandika Putra Wieke Wido Wati

Universitas Islam Negeri Fatmawati Sukarno Bengkulu,  STIESNU Bengkulu

Corresponding email: pebriputra@mail.uinfasbengkulu.ac.id

This study investigates the contrastive differences between English and Indonesian in expressing past tense, commonly referred to in Indonesian linguistics as the kala (tense) system. It aims to identify, describe, and analyze how past tense is marked in English compared to Indonesian, as well as to examine the translation techniques applied at the word, phrase, clause, and sentence levels. A descriptive qualitative approach with content analysis was employed, using R.E. Elson’s political biography of Suharto as the primary data source. The findings reveal that English follows strict morphosyntactic rules in constructing past tense sentences, whereas Indonesian exhibits no obligatory grammatical marking for tense, relying instead on contextual and lexical time markers. Regarding translation strategies, literal translation was the most frequently applied technique (24 data, 52%), followed by modulation (8 data, 17%), adaptation (3 data, 6.5%), and a combined category of alteration, omission, equivalence, and transposition (4 data, 8.6%). These findings underscore the structural asymmetry between the two languages in encoding temporal meaning and highlight the translator’s role in compensating for this asymmetry through diverse translation strategies. The study contributes to contrastive linguistics and translation studies by offering empirical evidence on how tense-related meaning is negotiated across typologically different languages.

Keywords: contrastive analysis; past tense; kala system; translation techniques; English–Indonesian translation 

STUDENTS’ AI USE DISCLOSURE IN EFL ACADEMIC WRITING: WHAT, WHY, AND HOW THEY REPORT AI ASSISTANCE 

Abstract ID: 150 

Noor Maulidiyah 

UIN Antasari Banjarmasin 

Corresponding email: noormaulidiyah@uin-antasari.ac.id 

This paper aims to analyze students’ AI use disclosures in their academic writing assignments. Specifically, it examines the generative AI tools students used, how they used them, and why they used them. It also explores the extent to which students transparently reported AI assistance in their writing. This study employed a qualitative descriptive research design. The data were collected from 49 students who completed a self-report on their use of AI tools while writing their EFL argumentative research papers. The students were given instructions on how to write the AI use disclosure. The findings reveal that ChatGPT was the most frequently used AI tool, primarily for brainstorming and generating ideas, organizing writing, and improving language. DeepL was predominantly used for translation. Although the students were instructed to elaborate on their AI use, many disclosures provided limited explanations of how and why the AI tools were used. The findings of this study contribute to the emerging literature on the responsible and ethical use of generative AI and academic integrity in EFL academic writing by highlighting AI-use disclosure as an important dimension of responsible AI practices and offering insights into how students communicate their AI-assisted writing processes.   

Keywords: AI Use Disclosure, Academic Writing, Transparency, Academic Integrity 

STUDENT ENGAGEMENT IN ARTIFICIAL INTELLIGENCE-BASED LEARNING: AN ANALYSIS OF BEHAVIORAL, EMOTIONAL, AND COGNITIVE ENGAGEMENT 

Abstract ID: 151 

Dyah Kusumastuti 

Universitas Muhammadiyah Purwokerto 

Corresponding email: dyahkusumastuti@ump.ac.id 

This study aims to describe the level of student engagement in Artificial Intelligence (AI)-based learning by examining three dimensions of engagement: behavioral, emotional, and cognitive. The respondents were students of the English Education Study Program at one of private universities in Banyumas, Central Java, Indonesia. Data were collected through a Likert-scale questionnaire (1–5) and analyzed descriptively. The results showed that student engagement was at a high level across all three dimensions. The average scores were 12.80 (out of 15) for behavioral engagement, 12.67 (out of 15) for emotional engagement, and 21.09 (out of 25) for cognitive engagement. These findings indicate students’ readiness to respond to the integration of AI in learning. Further research is recommended to analyze these aspects using a correlational quantitative approach or causal models such as Structural Equation Modeling (SEM) to explore the relationships between engagement and a broader range of factors. 

Keywords: Student Engagement, AI-based Learning, Higher Education, Artificial Intelligence

INVESTIGATING CHATGPT INTEGRATION IN ACADEMIC WRITING 

Abstract ID: 152 

Maria Olga Jelimun 

Universitas Katolik Indonesia Santu Paulus Ruteng 

Corresponding email: nimaria721@gmail.com 

This study aimed to examine the effectiveness of ChatGPT in improving the academic writing performance of first-year students in the English Language Education Study Program at Universitas Katolik Indonesia Santu Paulus Ruteng. As artificial intelligence continues to gain prominence in education, it is important to evaluate the effectiveness of ChatGPT as a tool to support English language learning, particularly in academic writing. This study employed a quantitative experimental research design. The participants were students from classes 2024 A, 2024 B, 2024 C, and 2024 D of the English Language Education Study Program. Total sampling was applied, involving all students from the selected classes. Data were collected through writing tests conducted under two conditions: without ChatGPT assistance and with ChatGPT assistance. A five-point Likert scale questionnaire was also administered to identify students’ perceptions of ChatGPT in academic writing. The data were analyzed using SPSS version 26 through descriptive statistics, normality and homogeneity tests, and a Paired Samples T-Test. The findings revealed a significant improvement in students’ writing performance after using ChatGPT. The mean score increased from 56.14 in the manual writing condition to 83.77 in the ChatGPT-assisted condition, with a significance value of 0.000 (p < 0.05). The results indicate that ChatGPT positively contributed to improving grammar accuracy, vocabulary usage, idea development, and organization. Furthermore, most students expressed positive perceptions toward the use of ChatGPT in academic writing. In conclusion, ChatGPT is an effective tool for enhancing students’ academic writing performance and can support English language learning when used critically and responsibly. 

Keywords: Chatgpt; Students’ Perception; Academic Writing; 

A DOUBLE-EDGED SWORD OF CHATGPT IN STUDENTS’ WRITING PROCESS HIGHLIGHTING BENEFITS AND POTENTIAL DANGERS 

Abstract ID: 157 

Ratih Novita Sari 

Universitas Muhammadiyah Prof Dr HAMKA 

Corresponding email: ratihnovitasari@uhamka.ac.id 

The integration of Artificial Intelligence (AI), particularly ChatGPT, into education presents both opportunities and challenges in the student writing process. This research explores the benefits and potential dangers of using ChatGPT for writing exposition texts among senior high school students. Using a qualitative approach with an Interpretative Phenomenological Analysis (IPA) design, data were collected from ten students of XI.2 Inglan 2 through in-depth interviews, observations, and document analysis. The findings reveal that ChatGPT provides significant support across all stages of writing. During pre-writing, it assists students in generating ideas and organizing outlines. In the drafting stage, it helps structure paragraphs effectively. During revision, ChatGPT improves grammar, vocabulary, and coherence. However, two major concerns emerged. First, excessive reliance on ChatGPT may lead to dependency, potentially weakening students’ critical thinking skills. Second, the use of AI may reduce writing originality, as students reported that their texts often became generic and lacked personal voice. This research concludes that ChatGPT can be a valuable tool in writing instruction when used responsibly. Educators should promote critical AI literacy to ensure that AI supports, rather than replaces, the development of students’ authentic writing abilities.

Keywords: ChatGPT, EFL Writing, Educational Technology, Student Dependency, Writing Originality 

HUMAN-AI CO-DESIGN IN ENGLISH MICROTEACHING: PRE-SERVICE TEACHERS’ REFLECTIONS ON LESSON SEQUENCING, PROMPTING, AND PEDAGOGICAL ALIGNMENT 

Abstract ID: 158 

Fika Megawati , Sheila Agustina, Saltanat Meiramova, Altyn Bakhitova

Universitas Muhammadiyah Sidoarjo

Corresponding email: fikamegawati@umsida.ac.id

This study explores how pre-service English teachers use generative artificial intelligence to support microteaching design, classroom practice, and lesson revision. The study responds to current debates in English language teaching that position AI as a pedagogical support tool requiring teacher agency, prompt literacy, and critical verification. Data were collected from open-ended reflections submitted after an English microteaching class. A qualitative descriptive design was used to examine students’ reflective accounts of AI-assisted lesson planning and teaching practice. The analysis followed open coding, category development, and thematic interpretation. The findings indicate that students mainly used AI to organize lesson sequences, develop teaching steps, design assessments, and simplify learning activities. They also reported difficulty in condensing core teaching activities into a short microteaching format while maintaining meaningful learning. In writing prompts, students tended to include topic, learner level, time allocation, learning model, and available media to obtain more contextual AI-generated suggestions. During microteaching practice, several challenges emerged, including time mismatch, passive learner responses, unclear classroom instructions, and the need to adjust activities to real classroom interaction. Students responded to these challenges by revising AI outputs, simplifying tasks, shortening instructions, and checking the alignment among objectives, activities, and assessment. The study concludes that AI can assist English microteaching design, but it cannot replace pedagogical judgment. AI-assisted microteaching should therefore emphasize prompt literacy, critical review, time testing, and objective-assessment alignment as core competencies in pre-service English teacher education. 

Keywords: generative AI, English microteaching, lesson planning, prompt literacy, AI literacy

AN INTEGRATED METACOGNITIVE-STRATEGIES FRAMEWORK FOR TECHNOLOGICAL-ENHANCED EFL LEARNING 

Abstract ID: 164 

Suciana Wijirahayu 

Universitas Muhammadiyah Prof. DR. HAMKA 

Corresponding email: sucianawijirahayu@uhamka.ac.id 

Purpose/Objective: The widespread adoption of digital technologies has profoundly transformed English as a Foreign Language (EFL) education, necessitating a deeper understanding of how learners effectively engage with these modern contexts. Drawing on Self-Regulated Learning (SRL) and Sociocultural Theories, this study investigates the crucial link between metacognition, learning strategies, and technology integration in EFL classrooms, aiming to understand how digital tools can move beyond mere availability to maximize educational impact. Methodology/Approach: To achieve this, we conducted a systematic literature review synthesizing existing research on digital EFL learning environments. This synthesis was utilized to develop a novel, comprehensive conceptual model: the Integrated Metacognitive-Strategic Framework for Technology-Enhanced EFL Learning. Key Findings/Outcomes: The review reveals that while technology significantly facilitates metacognitive processes and supports diverse learning strategies (cognitive, socio-affective, and compensatory), the development of these skills in digital environments is not automatic. Instead, successful integration heavily depends on intentional pedagogical design and explicit teacher support. Significance: These findings underscore that technology alone is insufficient for language acquisition; rather, the field of English language education must prioritize cultivating self-regulated, strategic digital learners. This framework provides educators and curriculum designers with an actionable approach to foster learners’ independence and enhance linguistic proficiency in an evolving digital landscape.

Keywords: Metacognition, Learning Strategies, EFL, Technology-Enhanced Learning, Self-Regulated Learning

“AI VS AI” LOOP IN HIGHER EDUCATION ELT: EXPLORING THE HUMAN ELEMENT IN AI-GENERATED ASSIGNMENTS AND STUDENT SUBMISSIONS 

Abstract ID: 214 

Nurteteng, Rizqi Claudia Wardani

Universitas Pendidikan Muhammadiyah Sorong

Corresponding email: nurteteng@unimudasorong.ac.id 

This qualitative case study investigates the emerging phenomenon of the “AI vs AI” loop in higher education ELT, where university lecturers utilize generative AI to design coursework and tasks while students simultaneously employ AI tools to generate their responses. Conducted at the English Language Education Department of Universitas Pendidikan Muhammadiyah Sorong, this study aimed to explore how the human element, specifically teacher agency, critical thinking, and student voice, is preserved or compromised within this automated cycle. Data were collected through semi-structured interviews with EFL lecturers and pre-service teacher students, document analysis of AI-generated assignments alongside student submissions, and student reflective logs. The findings reveal that relying solely on automated task generation and completion creates a mechanical learning environment that threatens academic integrity and diminishes student ownership over their L2 writing. However, when lecturers actively exercise pedagogical agency by redesigning tasks into process-based, context-rich, and reflective prompts, the automated cycle is effectively disrupted. Students were encouraged to critically evaluate AI outputs, integrating localized contexts and personal insights, which restored their critical agency and authentic voice. This study underscores the critical need to shift from passive AI adoption to human-centered ELT material design. The implications highlight that humanizing language education in the AI era relies not on banning generative tools, but on empowering educators to design AI-resilient, process-oriented tasks that cultivate higher-order thinking and preserve genuine human engagement in university EFL classrooms. 

Keywords: AI vs AI loop, teacher agency, student ownership, humanizing ELT, qualitative case study

BEYOND SURFACE CULTURE: RECONCEPTUALISING INTERCULTURAL COMMUNICATIVE COMPETENCE THROUGH GRAMMATICAL WORLDVIEW AWARENESS IN INDONESIAN EFL CONTEXT 

Abstract ID: 166 

Susanto, Pradnya Permanasari

Universitas Pekalongan

Corresponding email: susanto.unikal@gmail.com 

Research on intercultural communicative competence (ICC) in Indonesian EFL contexts has long conflated “culture” with its surface manifestations — customs, social values, and communicative practices — while systematically neglecting a deeper dimension: the way grammatical structures encode and reproduce distinct cultural worldviews. This conceptual article proposes a reconceptualisation of ICC that extends Byram’s (1997) foundational model by introducing grammatical worldview awareness (GWA) as a missing component of intercultural competence in EFL pedagogy. By elaborating on the relationship between the Indonesian–English agentivity divergence — as mapped in its typological predecessor — and the speaker’s self-construal as an agent in everyday events, this article argues that Indonesian EFL learners’ difficulty in producing natural English agentive constructions is not merely a grammatical problem, but reflects a deeper cultural unawareness of how language mediates the human relationship with events in the world. The article proposes a three-layered pedagogical framework — linguistic awareness, cultural reflection, and pragmatic production — intended to guide the development of EFL instructional materials that integrate ICC through grammatical consciousness-raising. The implications of this reconceptualisation are particularly relevant for EFL curriculum design at the senior high school level and for in-service EFL teacher training within the Indonesian Kurikulum Merdeka policy framework. 

Keywords: intercultural communicative competence, grammatical worldview awareness, Indonesian EFL, agentivity, ICC reconceptualisation, Kurikulum Merdeka 

EVALUATING THE ACCURACY OF NLP-BASED ERROR DETECTION AND AUTOMATED FEEDBACK IN ENGLISH STUDENT WRITING 

Abstract ID: 262 

Masrul, Safni Marwa, Fauzan Azim

Universitas Pahlawan Tuanku Tambusai

Corresponding email: masrulm25@gmail.com 

This study aims to evaluate the performance of a Natural Language Processing (NLP)-based system in detecting, correcting, and providing automated feedback on English writing errors made by English as a Foreign Language (EFL) students. This research used a machine learning-based experimental approach with a dataset consisting of 1,000 essays, which were divided into a 70% training set, 15% validation set, and 15% testing set. The system integrated BERT-based Error Detection, Transformer-based Grammar Correction, Semantic Similarity Analysis, and a Feedback Generation Module to identify five error categories, namely grammar, spelling, punctuation, word choice, and sentence structure errors. The evaluation results showed that this model achieved a macro F1 score of 0.9119, an AUC-ROC of 0.889, and an Average precision of 0.947. The highest performance was found in spelling errors with an F1 score of 0.939, while sentence structure errors showed relatively lower performance. In addition, this system obtained a semantic similarity score of 0.930 and a human evaluation of 4.40/5, which indicates that the correction results were able to maintain the meaning of the writing and the feedback was considered quite accurate, relevant, and clear. Overall, the findings of this study indicate that the integration of NLP and transformer-based models has the potential to support the development of automated feedback capable of assisting students and teachers in the process of evaluating and revising English writing more quickly and systematically. 

Keywords: Automated Feedback, EFL Writing, Error Detection, Grammar Correction, Natural Language Processing. 

AUTHENTICITY JUDGMENTS OF INDONESIAN EFL STUDENTS IN GENERATIVE AI-ASSISTED ENGLISH LEARNING: AN INTERPRETIVE QUALITATIVE STUDY

Abstract ID: 274 

Selamat Husni Hasibuan, Heridayani, Pirman Ginting, Rini Susanti

Universitas Muhammadiyah Sumatera Utara

Corresponding email: selamathusni@umsu.ac.id 

This study examines how undergraduate EFL students conceptualise authentic learning when generative artificial intelligence (AI) becomes part of their English learning practices. It addresses the question of how students determine whether AI-assisted activities still constitute genuine learning. An interpretive qualitative design was employed at a private university in Medan, North Sumatra, Indonesia. Six undergraduate EFL students who had experience using generative AI for English learning participated in semi-structured interviews conducted between March and June 2026. Interviews were conducted individually in Bahasa Indonesia, audio-recorded with informed consent, transcribed verbatim, and translated into English. Data were analysed inductively using reflexive thematic analysis following Braun and Clarke’s six-phase framework, with attention to participants’ experiences, reflections, and judgments concerning AI assistance, cognitive effort, ownership, and learning. The findings indicate that students valued AI for its accessibility, immediate explanations, and ability to facilitate understanding. However, they also raised concerns about reduced intellectual effort, cognitive dependence, and delegating thinking. Authentic learning was not defined by the absence of AI but by meaningful cognitive engagement, independent judgment, personal understanding, and responsibility for learning. The study suggests that authentic learning in AI-mediated EFL education is an ongoing process of authenticity judgment, through which learners negotiate the boundary between meaningful assistance and cognitive substitution. These findings highlight the need for English language educators to move beyond simply regulating AI use and instead develop students’ capacity to make informed judgments about when and how AI supports genuine learning.

Keywords : generative artificial intelligence, authentic learning, authenticity judgments, AI-assisted language learning, EFL students.

ADAPTIVE AI IN EFL HIGHER EDUCATION: FROM COGNITIVE PERSONALIZATION TO AFFECTIVE SUPPORT 

Abstract ID: 275

Sulkifli

Universitas Muhammadiyah Makassar

Corresponding email: sulkifli@unismuh.ac.id 

Artificial intelligence is increasingly used in English as a Foreign Language (EFL) education to personalize content, task difficulty, pace, and feedback. Yet personalization is still commonly driven by performance data, while the affective conditions of language learning receive less systematic attention. This conceptual paper examines how adaptive AI can retain cognitive personalization while incorporating pedagogically responsible affective support. It develops a theory-building synthesis from verified foundational work on foreign language anxiety, research on achievement emotions, recent reviews of personalized adaptive learning and AI-assisted EFL emotions, empirical studies published through 2026, and UNESCO guidance on human-centred AI. The evidence is mixed rather than uniformly positive: AI-mediated interaction can create low-risk practice, flexible pacing, and personalized feedback, but effects on anxiety vary by task and design, and poorly calibrated feedback or opaque emotion recognition can introduce new pressure. Drawing these strands together, the paper proposes a dual-layer framework in which cognitive personalization is complemented by five affective-support functions: calibrated challenge, low-stakes rehearsal, feedback regulation, learner agency and pacing, and human oversight with data restraint. Affective support is treated as a pedagogical design principle, not a clinical mental-health function. The framework offers a practical basis for designing and testing adaptive AI that responds not only to what EFL learners know, but also to the conditions under which they can participate, take communicative risks, and learn. 

Keywords: Adaptive AI; affective support; cognitive personalization; EFL; foreign language anxiety

CHATGPT-BASED INTELLIGENT INSTRUCTIONAL MODEL FOR ENHANCING EFL LEARNERS’ WRITING SKILLS: STUDENTS AND TEACHERS’ PERCEPTIONS AND PRACTICES 

Abstract ID: 209 

Ilham 

Universitas Muhammadiyah Mataram 

Corresponding email: ilham.ummataram@gmail.com 

Chat GPT is currently receiving a lot of attention particularly in writing especially in writing and globally. Writing in higher education has been the subject of numerous studies. Chat GPT is a new technological tool that many students are using.  With regard to technological advancements, artificial intelligence (AI)-based technologies like chat GPT can help  overcome difficulties with paragraph writing. However, teachers’ and students’ perception of Chat GPT as a digital tool for English writing are known. Using a qualitative case study approach, the study explores lecturers and students’ perspective on using Chat GPT as a tool to support English writing. Three lecturers who were familiar with chat GPT and nine students participated. Questionnaires and in-depth interviews are the instruments to collect the data. Findings showed students and teachers responded positively overall to using Chat GPT to help improve students’ writing quality. Additionally, Chat GPT was providing useful writing features, and aiding language development. This study suggests AI technologies like Chat GPT play an important role in facilitating high-quality student paragraph writing.

Keywords: Writing; artificial intelligence (AI); chat GPT; EFL students and teacher

MAKING LEARNING VISIBLE IN THE AI ERA: EXPLORING AI-ASSISTED LESSON-PLAN REVISION AMONG HIGH-PERFORMING PRE-SERVICE EFL TEACHERS

Abstract ID: 277

Dewi Atikah, Abd Haris Takulani, Nur Hasanah Safei, Anita Rezki

UIN Kendari

Corresponding email: dewi.atikah70@gmail.com 

Generative artificial intelligence (AI) has increasingly transformed how pre-service teachers develop instructional materials, raising an important assessment question: how can teacher educators evaluate student learning when AI can generate seemingly competent lesson plans? This qualitative multiple-case study explores how high-performing pre-service EFL teachers use, evaluate, and transform AI-generated pedagogical suggestions through peer and lecturer feedback. Two pre-service teachers, selected as the first- and third-place winners of a microteaching competition, participated in the study. Data consisted of participants’ AI prompt and response records, initial and revised lesson plans, peer and lecturer feedback, written reflections, and semi-structured interviews. Data were analyzed thematically, with particular attention to AI use, feedback engagement, revision decisions, pedagogical reasoning, and evidence of learning. The study focuses on revision episodes in which participants encountered AI suggestions and human feedback, evaluated their relevance, and decided whether to accept, modify, or reject them. Preliminary analysis indicates that AI functioned not only as a content generator but also as a source of ideas and pedagogical support. However, participants did not consistently accept AI-generated suggestions, particularly when these conflicted with considerations of learner level, time, feasibility, and instructional objectives. Peer and lecturer feedback prompted further reflection on alignment among learning objectives, activities, assessment, and learner needs. The findings suggest that students’ revision processes and reflective justifications provide valuable evidence of pedagogical learning that cannot be captured by evaluating the final AI-assisted lesson plan alone. The study highlights the potential of process-oriented assessment for making learning visible in AI-mediated teacher education. 

Keywords: Generative AI, EFL Teacher Education, Lesson Planning, Process-Oriented Assessment, Pedagogical Learning

HUMAN-CENTERED ASSESSMENT OF DIGITAL ENGLISH FOR BUSINESS COMMUNICATION IN THE AGE OF AI 

Abstract ID: 276 

Ali Syahban Amir

Institut Turatea Indonesia

Corresponding email: alisyahbanamir.intijpt@gmail.com 

Artificial intelligence is reshaping language assessment, yet digitally mediated instruction still requires human judgment, contextualized tasks, and transparent criteria. This study examined the effectiveness and assessment implications of digital instructional materials for English for Business Communication (EBC) among undergraduates in a Development Economics program at an Indonesian higher education institution. Using the implementation and evaluation phases of an ADDIE-based research-and-development study, 30 students participated in small- and larger-group trials. Data were obtained from pre- and post-tests, observations, student and lecturer response questionnaires, and analytic rubrics covering vocabulary, email coherence, and business language use. Mean performance increased from 62.4 to 80.3, representing a 17.9-point gain (t = 6.85, p < .001). Vocabulary improved from 68.4 to 79.2, email coherence from 65.0 to 76.5, and business language use from 61.2 to 74.3, with the latter showing the largest proportional increase. Student and lecturer responses reached 89.6% and 91.2%, respectively, indicating strong acceptance and practicality. The findings show that digital EBC instruction can support measurable learning gains when assessment remains authentic, multidimensional, and lecturer-mediated. In the age of AI, technology should function as a scaffold for practice and reflection rather than replace human evaluation. The study supports assessment designs that combine digital access, authentic professional communication, analytic rubrics, and teacher judgment to preserve both rigor and human agency

HUMANIZING PRONUNCIATION PRACTICE: INTEGRATING AI-ASSISTED SPEECH TOOLS WITH PHONETIC DRILLING TO ENHANCE STUDENTS’ SPEAKING SKILLS 

Abstract ID: 281

Iim Khairunnisa, Dalpah Triwidiah, Eka Dewi Fithrotunnisa, Putri Silvia, Aida Afrilia

STKIP Syekh Manshur

Corresponding email: iimkhairunnisa89@gmail.com

In many English classrooms, helping students speak with confidence and clear pronunciation remains a daily challenge, often due to limited practice time and anxiety over making mistakes. This study explores a warm, human-centered approach to blending modern technology with traditional language teaching by integrating AI-assisted speech tools into the phonetic drilling method. Rather than replacing the personal touch of a classroom, digital speech tools serve as a supportive companion that gives students a safe, private space to practice their pronunciation at their own pace. Using a true experimental design, the research involved sixty eleventh-grade students at MAN 3 Pandeglang, divided equally into an experimental group experiencing this blended approach and a control group. The quantitative findings reveal a meaningful improvement in the students’ speaking proficiency and clarity. Beyond the numbers, the classroom observations highlight a wonderful shift: when technology handles repetitive pronunciation drills, teachers have more emotional space and time to mentor, encourage, and connect deeply with their students. Ultimately, this study shows how thoughtful, humanized technology integration can breathe new life into English language teaching, making speaking practice both effective and deeply personal.

Keywords: Phonetic drilling, speaking skills, humanizing technology, English language teaching, pronunciation

WHEN AI MEETS TORAJA: HUMANIZING ENGLISH LANGUAGE EDUCATION THROUGH CULTURE, ETHICS, AND TEACHER AGENCY

Abstract ID: –

Markus Deli Girik Allo

Toraja Christian University, Indonesia

Corresponding email: –

Generative artificial intelligence (GenAI) is reshaping English language education by enabling rapid text generation, feedback, translation, lesson planning, and learning-material development. Yet the educational value of these tools cannot be separated from the cultural and relational contexts in which language teaching occurs. This study investigates how a human-centred approach to AI can be conceptualized for English language education in Toraja, Indonesia, with particular attention to culture, ethics, and teacher agency. A qualitative document analysis was conducted on 11 purposively selected sources published between 2023 and 2026, comprising international AI-in-education guidance, empirical and conceptual studies on GenAI in language education and teacher agency, and a Toraja-specific study of high-context EFL classroom communication. The documents were analyzed through iterative coding, cross-document comparison, and thematic synthesis. Four interrelated themes were identified: (1) AI is most educationally defensible when it augments rather than displaces human pedagogical judgment; (2) culture functions as an interpretive filter for AI-generated language and representations; (3) AI ethics in language education is relational as well as procedural, involving authorship, privacy, accuracy, cultural representation, and responsibility; and (4) teacher agency operates as a form of pedagogical governance that determines when, why, and how AI should be used. From these findings, the study develops the Culture-Ethics-Teacher Agency (CETA) Framework for Humanizing AI in English Language Education. The framework positions AI as a pedagogical resource under human and culturally situated oversight. The study concludes that responsible AI integration in Toraja should be evaluated not by the amount of technology used, but by its contribution to meaningful learning, cultural integrity, ethical responsibility, and sustained teacher-student relationality.

Keywords: generative artificial intelligence; English language education; Toraja; high-context culture; ethics; teacher agency; human-centred pedagogy

THE ANALYSIS OF STUDENTS PERCEPTIONS OF SONGS, PODCASTS, AND LEARNING STRATEGIES IN LISTENING SKILLS DEVELOPMENT 

Abstract ID: 270 

Intan Suryani Shaputri 

Universitas Muhammadiyah Semarang 

Corresponding email: putrisha234@gmail.com 

Listening plays a crucial role in mastering English as a foreign language because it supports comprehension and interaction, which are essential for successful communication. For many learners, listening forms the first gateway to understanding English input before they are able to produce speech, making it one of the most fundamental language skills. However, students frequently find listening difficult due to factors such as unfamiliar vocabulary, fast speech, and limited exposure to authentic materials. The purpose of this study is to explore students’ perceptions of improving listening skills through songs, podcasts, and metacognitive learning strategies. This research was conducted at Universitas Muhammadiyah Semarang by distributing a ten-item questionnaire to 30 students of the English Education Department. The questionnaire assessed their opinions regarding the effectiveness of songs, podcasts, and strategies such as planning, monitoring, and evaluating in enhancing their listening ability. The findings revealed that students generally viewed listening as a challenging skill but believed that digital media and self-regulated learning strategies could make the process more enjoyable and effective. Integrating songs, podcasts, and metacognitive strategies was found to enhance motivation, focus, and comprehension. The study concludes that technology-based learning combined with reflective learning techniques can help EFL students build confidence and autonomy in listening comprehension. In addition to its role in language learning, listening also plays a significant function in students’ academic success, particularly in higher education contexts. In the context of higher education, listening skills are not only required in language classrooms but also across academic disciplines. Many learning resources, such as online lectures, academic presentations, webinars, and instructional videos, are delivered in English. As a result, students with limited listening proficiency may experience difficulties understanding course materials, participating in discussions, and completing academic tasks. Therefore, improving listening skills becomes a crucial academic requirement rather than merely a language learning objective. Along with these academic demands, the increasing use of digital technology in education has changed how students access and engage with learning materials. Today’s students are exposed to various forms of digital audio content through smartphones and personal devices. This condition creates opportunities for integrating digital media, such as songs and podcasts, into language learning. When listening materials are aligned with students’ daily habits, learning becomes more meaningful and relevant. Consequently, understanding students’ perceptions of digital listening practices is essential for designing effective and contextualized instruction.

Keywords: listening skill, songs, podcasts, metacognitive strategy, students’ perception

THE IMPORTANCE, CHALLENGES, AND STRATEGIES IN DEVELOPING ENGLISH LISTENING COMPREHENSION 

Abstract ID: 265 

Hana Ibtihaj 

Universitas Muhammadiyah Semarang 

Corresponding email: hanaibti05@gmail.com 

In learning English, listening comprehension is essential. However, in many classrooms, it is still the most neglected skill to be developed. It is the foundation skill for all the other language skills (speaking, reading, and writing) to develop. It is still a significant means of alienation and communication in everyday life. The process of listening in class is quite the opposite of the communication skills and topics that are mainly concentrated on in the class. Listening is one of the most critical aspects in language learning, but its practice is still limited. A gap has been found between teaching and implementing listening skills in learners, and studies confirm that this gap affects positively or negatively the learners’ ability to handle real conversation. Pointing to the many reasons for the struggles of the EFL students, research mentions several issues like the speed of natural speech, accents, and others. Besides, some features of connected speech make it even more difficult. Before all that, psychological issues (anxiety, for instance) are at work. The good thing is that there is some evidence that strategies can help one overcome these difficulties. Normally, the way that conventional practices take up the areas of language comprehension and skill building, with the former being the priority, raises serious questions in language education. The perfect idea is to use the Task Based Language Teaching approach. The use of multimedia tools and technology that expose students to authentic and exciting material is another central point. Trainers can teach cognitive and metacognitive strategies promoting independence among students. It appears that going for the interactive and learner-centered methods can change the scenario a lot. The changes made would not only make the students strong listeners but also confident users of spoken English. 

Keywords: Listening Comprehension, English Language Learning, EFL Challenges, Teaching Strategies. 

THE ANALYSIS OF INNOVATIVE DIGITAL APPROACHES TO LISTENING SKILLS IN EFL CONTEXTS 

Abstract ID: 267 

Safira Putri Aprillia 

Universitas Muhammadiyah Semarang 

Corresponding email: sfiratssss@gmail.com 

Listening skills are a crucial foundation in English as a Foreign Language (EFL) learning, yet they are often neglected due to limited exposure to authentic spoken language and passive teaching methods. To address these challenges, various studies have focused on teaching innovations and technological interventions aimed at improving learners’ listening comprehension. Research indicates that the integration of digital technology plays a key role in this development. Approaches such as the use of mobile learning applications provide flexible and interactive learning experiences, while digital audio platforms such as podcasts and music streaming services have also been found to strengthen learners’ listening proficiency through repeated and self paced exposure to authentic speech (Angelia & Simanjuntak, 2023; Salainti & Pratiwi, 2021). Similarly, the utilization of social media platforms like Instagram has been shown to boost motivation and exposure to authentic listening practices through English videos and content (Agustin & Ayu, 2021), and digital storytelling has likewise been found to significantly improve EFL learners’ listening comprehension and attitudes (Akdamar & Sütçü, 2021). In addition to technology, creative pedagogical strategies are also important. The use of games (such as Running Dictation and Whispering Games) creates an enjoyable learning atmosphere, reduces anxiety, and promotes active participation (Syafii et al., 2020). Furthermore, formative online assessment via Learning Management Systems (LMS) is highlighted for providing instant feedback, objectivity, and continuous support for learners (Slamet & Mukminatien, 2024). Overall, these studies point to a paradigm shift from conventional methods towards interactive and technology based learning, confirming that digital tools (mobile apps, social media, podcasts, LMS) and strategies like games are highly effective in improving the listening competence of modern learners. This study employs a quantitative research design. The main instrument is a structured questionnaire delivered via Google Forms. The questionnaire is closed-ended with 15 statements, measured using a three-point Likert scale (1 = agree, 2 = neutral, 3 = disagree) to obtain standardized ordinal data and minimize interpretation bias. 

Keywords: Listening Skills, EFL, Technology Based Learning, Mobile Learning, Games, Quantitative Research. 

EFL STUDENTS’ PERCEPTIONS OF MOBILE LEARNING AS A RESOURCE FOR SPEAKING CLASSES 

Abstract ID: 263 

Nanda Fatika Khasanah 

University of Muhammadiyah Semarang 

Corresponding email: nandafatika457@gmail.com 

This study explores EFL students’ perceptions of mobile learning as a resource for speaking classes, focusing on perceived learning benefits, motivation, interaction, comfort, and overall satisfaction. Speaking is often challenging for EFL learners because opportunities to practice are limited and students may experience anxiety when speaking in conventional classrooms. Mobile learning can provide opportunities for students to practice speaking beyond classroom hours. This study used a quantitative approach with a structured questionnaire based on a five-point Likert scale. The questionnaire was administered to EFL students who had received speaking instruction via mobile-based platforms, including WhatsApp, Google Classroom, and similar applications. The data were analyzed descriptively to identify patterns in students’ responses. The findings show that students viewed mobile learning as a useful resource for speaking practice. It provided flexibility, encouraged participation, and helped reduce feelings of shyness and fear. However, students also reported several difficulties, particularly unstable internet connections and limited teacher feedback. Many students still preferred face-to-face speaking classes because they felt more motivated and received direct interaction. Therefore, mobile learning is more suitable as a complementary resource within blended learning. 

Keywords: mobile learning, EFL speaking, students’ perceptions, speaking classes, blended learning.

COMMUNICATION INTERCHANGE: INVESTIGATING SPEECH ACCOMMODATION IN LECTURER-LEARNERS INTERACTION IN ELT AT UNIVERSITAS MUHAMMADIYAH MAKASSAR 

Abstract ID: 246 

Adni Dzilarsy Makina, Eny Syatriana, Ika Sastrawati

Universitas Muhammadiyah Makassar

Corresponding email: adni@bg.unismuhmakassar.ac.id 

Communication in English Language Teaching (ELT) involves continuous interaction between lecturers and learners, in which participants may adjust their communication according to the interactional context. This study investigates the forms of communication accommodation employed by a lecturer and learners during classroom interaction in an ELT context at Universitas Muhammadiyah Makassar. A qualitative descriptive design was employed to examine naturally occurring communication in a Speaking for Advanced class. Data were collected through classroom observations across five instructional meetings and semi-structured interviews with selected students. The data were analyzed using Communication Accommodation Theory (CAT), particularly the concepts of convergence, divergence, and maintenance (Giles, 1973). The findings identified all three forms of communication accommodation. Convergence was the most prominent form and occurred through adjustments in vocabulary, speech rate, pronunciation, sentence structure, repetition, and language choice by both the lecturer and learners. Divergence appeared less prominently when participants maintained communicative differences, including target-language expectations and certain established learner communication patterns. Maintenance occurred when participants continued established language choices, response patterns, or interactional styles without substantial adjustment. Overall, the findings indicate that communication accommodation was reciprocal but not necessarily symmetrical, as both the lecturer and learners adjusted their communication while maintaining aspects of their established communicative patterns. The study highlights the context-dependent nature of communication accommodation and demonstrates how both lecturers and learners actively shape classroom interaction in ELT. 

Keywords: classroom interaction; communication accommodation; convergence; divergence; English Language Teaching 

BEYOND COPY-PASTE: REGULATING GENERATIVE AI VIA PROMPT GUIDEBOOK AND ITS IMPACT ON EFL STUDENTS’ WRITING EFFICACY AND RHETORICAL QUALITY 

Abstract ID: 253 

Andi Nur Alfiah Insani Amin, Ismail Sangkala, St. Asmayanti AM

Universitas Muhammadiyah Makassar

Corresponding email: andinuralfiah@bg.unismuhmakassar.ac.id 

Generative AI integration in EFL writing instruction has gained significant attention, yet previous studies have largely overlooked the ethical process of how students interact with these tools. This quasi-experimental research investigates the effectiveness of the DeepSeek application as a regulated digital scaffolding tool in improving students’ writing skills in composing hortatory exposition texts. Employing a non-equivalent control group design, the study involved 50 eleventh-grade students divided into experimental and control groups, with data analyzed through descriptive statistics, normality and homogeneity tests, Wilcoxon test, and Mann-Whitney U test. The findings reveal that both DeepSeek-assisted instruction and conventional methods significantly improved students’ writing scores, as evidenced by the Wilcoxon test results (p = 0.000). Strikingly, the Mann-Whitney U test revealed no statistically significant difference between the two groups (p = 0.748 > 0.05), a pedagogical breakthrough demonstrating that ethically-guided AI use, strictly regulated by a Prompt Guidebook, produces outcomes comparable to well-designed human collaborative learning. This parity confirms that students maintained cognitive agency throughout the writing process, proving that teachers and effective pedagogy remain irreplaceable in the digital age. These findings carry profound implications for EFL pedagogy, curriculum development, and institutional AI policies. 

Keywords: DeepSeek Application, Writing Skills, Hortatory Exposition Text

NAVIGATING ACADEMIC WRITING WITH AI: HOW STUDENTS EMPLOY AI WRITING TOOLS ACROSS THE WRITING PROCESS 

Abstract ID: 252 

Regia Mutmainna Zyam, Ardiana, Muhammad Zia Ul Haq

Universitas Muhammadiyah Makassar

Corresponding email: regia@bg.unismuhmakassar.ac.id 

Academic writing is demanding for EFL undergraduate students, who must plan, draft, and revise texts. AI writing tools offer immediate feedback and language support, making them a potential aid throughout this process. This study examines how students employ these tools for different purposes during the planning, drafting, and revising stages of academic writing. A qualitative descriptive-exploratory design was used, involving ten purposively selected English Education students with prior experience using AI writing tools. Data were collected through semi-structured interviews and non-participant observation, then analyzed according to each stage of the writing process. The findings show that students used AI throughout the entire writing process, but for different purposes at each stage: during planning, AI helped clarify task instructions and generate topic ideas, while topic selection remained the student’s own decision; during drafting, AI was mainly used to expand and connect ideas after an initial draft had been written, though this order sometimes shifted under deadline pressure; and during revising, AI was used to check content flow and correct grammar and vocabulary, with students applying their own judgment in deciding which suggestions to accept. These findings indicate that AI functioned as a recursive support tool integrated across the writing process rather than replacing any single stage, while students retained control over key writing decisions. This study contributes to a deeper understanding of AI-assisted academic writing practices and offers pedagogical implications for English language education. 

Keywords: AI writing tools, academic writing, writing process, EFL students

AN INVESTIGATION OF CRITICAL THINKING IN PERSUASIVE WRITING OF MALE AND FEMALE NINTH GRADE STUDENTS 

Abstract ID: 258 

Siti Munhafiroh

S2 Pascasarjana universitas Islam Negeri Raden Mas Said Surakarta

Corresponding email: smunhafiroh@gmail.com 

Critical thinking is essential in persuasive writing, requiring students to articulate opinions, justify arguments, and evaluate complex issues. This research investigated critical thinking skills in persuasive writing among ninth-grade students at SMP Al Firdaus Surakarta during the 2025/2026 academic year, while also examining gender-based differences. Grounded in Facione’s (1990) framework, the research analyzed six core skills: interpretation, analysis, evaluation, inference, explanation, and self-regulation. Employing a qualitative case research design, six students (three males, three females) were purposively selected. Data were gathered via classroom observations, semi-structured interviews, and document analysis of 12 persuasive essays (194 sentences) covering two topics: the Free Nutritious Meal Program (MBG) and Mobile Phone Use at School. The findings revealed that students demonstrated all six skills, with explanation being the most prevalent, followed by self-regulation, evaluation, analysis, inference, and interpretation (the least frequent). Most skills were displayed at a basic level, categorized overall as “developing critical thinkers.” Familiarity with the mobile phone topic generated richer arguments than the MBG topic. While both genders exhibited all six skills, male students presented arguments more directly, whereas female students offered more detailed, reflective explanations. Explicit instruction remains vital to further elevate these skills. 

Keywords: Critical Thinking; Persuasive Writing; EFL Students; Facione Framework 

IMPROVING STUDENTS’ READING COMPREHENSION USING MEDIUM APPLICATION 

Abstract ID: 245 

Andi Fahima Yanti, Ummi Khaerati Syam, Ariana

Universitas Muhammadiyah Makassar

Corresponding email: andifahima@bg.unismuhmakassar.ac.id 

Reading comprehension remains a persistent challenge for Indonesian EFL students, particularly at the inferential level, which requires readers to draw conclusions and understand implied meaning beyond explicit text content. This study aimed to explore the impact of implementing the Medium application, an open publishing platform emphasizing authentic long-form content, on students’ reading comprehension in narrative texts, and to examine the factors contributing to students’ reading difficulties. The research employed a Classroom Action Research (CAR) design based on the Kemmis and McTaggart model, conducted over two cycles with 38 grade XI students at SMAN 18 Makassar, selected through purposive sampling. Data were collected through diagnostic and cycle-based reading comprehension tests covering five indicators: finding the main idea, identifying detailed information, identifying text structure, understanding implied meaning, and drawing conclusion, and were analyzed using descriptive statistics in SPSS. The results showed a significant and consistent improvement in students’ mean scores, from 58.75 in the diagnostic test to 83.03 in Cycle I and 91.71 in Cycle II, with the proportion of students in the Excellent category rising from 31.6% to 68.4%. The most notable gain occurred in identifying text structure, following explicit instruction introduced in Cycle II. These findings indicate that Medium, when integrated with structured pedagogical guidance, can serve as an effective authentic digital reading platform for strengthening students’ digital literacy and improving inferential reading comprehension in the EFL classroom. 

Keywords: classroom action research, digital reading platform, inferential comprehension, Medium application, reading comprehension 

TRANSLANGUAGING PRACTICES AND COMMUNICATIVE FUNCTIONS IN RELIGIOUS-ENGLISH LANGUAGE EDUCATION CONTENT ON TIKTOK: IMPLICATIONS FOR DIGITAL LITERACY 

Abstract ID: 188 

Amalia Kartika, Sulastri

Universitas Muslim Indonesia

Corresponding email: amelialia9309@gmail.com 

Translanguaging is an approach that views multilingual speakers as users who integrate their entire linguistic repertoire to construct meaning (García & Wei, 2014). This study aims to identify translanguaging practices in religious and English language education content on TikTok, to examine the communicative functions of translanguaging across both types of content, and to explore the implications of these translanguaging practices for digital religious–English language education content. The study focused on two genres of TikTok content in Indonesia: English language learning videos by @NakulaEnglish and religious sermon (ceramah) content by @Okisetianadewi13, both combining Indonesian, Arabic, and English. The dataset consisted of ten TikTok videos, five from each account. The analysis was grounded in the theory of translanguaging (García & Wei, 2014) and pedagogical translanguaging (Cenoz & Gorter, 2021, 2022). Using a qualitative approach involving content and discourse analysis of TikTok transcripts, the study identified patterns of language use and their communicative functions. The findings showed that both accounts employed the same five forms of translanguaging but with markedly different distributions: @NakulaEnglish relied more on insertion and intrasentential translanguaging for pedagogical purposes, while @Okisetianadewi13 predominantly used entire translanguaging to preserve religious authority and accommodate diverse audiences. The results indicated that translanguaging functioned not only as a linguistic strategy but also as a pedagogical and communicative resource for enhancing audience understanding, building rapport, strengthening speaker authority, and shaping hybrid digital identities. The findings have important implications for the development of multilingual and culturally responsive digital learning environments. 

Keywords: Translanguaging; Pedagogical Translanguaging; TikTok; religious discourse; English language education

DISCURSIVE ENACTMENTS OF DIGITAL TEACHER AGENCY IN ENGLISH MICROLEARNING: A QUALITATIVE CONTENT ANALYSIS OF INSTRUCTIONAL DISCOURSE IN INSTAGRAM REELS 

Abstract ID: 230 

Nurhalizah Bachtiar 

Universitas Muslim Indonesia 

Corresponding email: nurhalizahzaaa@gmail.com 

Instagram Reels have become a prominent space for short-form English instruction, yet limited attention has been given to how digital educators enact teacher agency through instructional discourse. This preliminary study examines how digital teacher agency is discursively enacted in English microlearning on Instagram. The data comprise 30 purposively selected transcripts from three public English-learning accounts, with ten Reels from each account. Using qualitative content analysis, teacher agency is operationalized as observable pedagogical choices and positioning manifested through explaining, exemplifying, contrasting, correcting, reformulating, modelling, and engaging viewers. An initial reading identifies three interconnected enactments of agency. First, curricular agency is reflected in the selection and simplification of language content for brief, focused lessons. Second, pedagogical agency is enacted through contextualized examples, corrective feedback, reformulated responses, humour, and scenario-based explanations. Third, interactional agency emerges through direct address, rhetorical questions, and invitations for audience participation. The accounts also display distinct instructional orientations, including vocabulary and idiom explanation, IELTS speaking coaching, and contextualized error correction. These preliminary patterns suggest that digital teacher agency in short-form English education is constructed not only through what educators teach but also through how they linguistically organize explanations and position viewers as learners. The study contributes to discussions of teacher agency and professional identity in digital English language teaching and offers implications for preparing prospective teachers to design concise, engaging, and pedagogically purposeful microlearning content. 

Keywords: digital teacher agency, English microlearning, instructional discourse, Instagram Reels, qualitative content analysis 

ENGAGING LEARNERS THROUGH DIGITAL TECHNOLOGY: THE EFL PRESERVICE TEACHERS AGENCY DURING PRACTICUM 

Abstract ID: 232 

Sri Lestari, Ahmad Bukhori Muslim, Ika Lestari Damayanti

Universitas Pendidikan Indonesia

Corresponding email: sri.lestari@upi.edu 

This study explored how EFL preservice teachers in Indonesia exercised their agency by integrating digital technology during a school-based teaching practicum. An ecological approach to teacher agency was employed in this study to examine the interactions among EFL preservice teachers’ beliefs about effective teaching, classroom realities, resources, and school constraints, and future aspirations. The study employed a qualitative case study design involving five EFL preservice teachers participating in a three-month teaching practicum at a high school. Data were collected through lesson plans, classroom observations, reflective journals, and interviews, and were analyzed thematically. The findings revealed that EFL preservice teachers actively initiated technology integration to enhance engagement, motivation, and relevance in their English lessons. They utilized digital tools to organize lessons, provide examples, support learning activities, encourage participation, and assess learning. However, their agency was often constrained by classroom conditions and rules. In response, they negotiated these challenges and developed alternative strategies. Additionally, the study found that EFL preservice teachers intend to continue integrating technology into their future teaching. This emphasizes the need for teacher education programs to prepare future English teachers to effectively integrate digital technology into their lessons, select appropriate tools for diverse contexts, and anticipate potential challenges.  

Keywords: digital technology, ecological approach, preservice teacher, teacher agency, teaching practicum

“AI VS AI” LOOP IN HIGHER EDUCATION ELT: EXPLORING THE HUMAN ELEMENT IN AI-GENERATED ASSIGNMENTS AND STUDENT SUBMISSIONS 

Abstract ID: 214 

Nurteteng, Rizqi Claudia Wardani

Universitas Pendidikan Muhammadiyah Sorong

Corresponding email: nurteteng@unimudasorong.ac.id 

This qualitative case study investigates the emerging phenomenon of the “AI vs AI” loop in higher education ELT, where university lecturers utilize generative AI to design coursework and tasks while students simultaneously employ AI tools to generate their responses. Conducted at the English Language Education Department of Universitas Pendidikan Muhammadiyah Sorong, this study aimed to explore how the human element, specifically teacher agency, critical thinking, and student voice, is preserved or compromised within this automated cycle. Data were collected through semi-structured interviews with EFL lecturers and pre-service teacher students, document analysis of AI-generated assignments alongside student submissions, and student reflective logs. The findings reveal that relying solely on automated task generation and completion creates a mechanical learning environment that threatens academic integrity and diminishes student ownership over their L2 writing. However, when lecturers actively exercise pedagogical agency by redesigning tasks into process-based, context-rich, and reflective prompts, the automated cycle is effectively disrupted. Students were encouraged to critically evaluate AI outputs, integrating localized contexts and personal insights, which restored their critical agency and authentic voice. This study underscores the critical need to shift from passive AI adoption to human-centered ELT material design. The implications highlight that humanizing language education in the AI era relies not on banning generative tools, but on empowering educators to design AI-resilient, process-oriented tasks that cultivate higher-order thinking and preserve genuine human engagement in university EFL classrooms. 

Keywords: AI vs AI loop, teacher agency, student ownership, humanizing ELT, qualitative case study

AN INTEGRATED METACOGNITIVE-STRATEGIES FRAMEWORK FOR TECHNOLOGICAL-ENHANCED EFL LEARNING 

Abstract ID: 164 

Suciana Wijirahayu 

Universitas Muhammadiyah Prof. DR. HAMKA 

Corresponding email: sucianawijirahayu@uhamka.ac.id 

Purpose/Objective: The widespread adoption of digital technologies has profoundly transformed English as a Foreign Language (EFL) education, necessitating a deeper understanding of how learners effectively engage with these modern contexts. Drawing on Self-Regulated Learning (SRL) and Sociocultural Theories, this study investigates the crucial link between metacognition, learning strategies, and technology integration in EFL classrooms, aiming to understand how digital tools can move beyond mere availability to maximize educational impact. Methodology/Approach: To achieve this, we conducted a systematic literature review synthesizing existing research on digital EFL learning environments. This synthesis was utilized to develop a novel, comprehensive conceptual model: the Integrated Metacognitive-Strategic Framework for Technology-Enhanced EFL Learning. Key Findings/Outcomes: The review reveals that while technology significantly facilitates metacognitive processes and supports diverse learning strategies (cognitive, socio-affective, and compensatory), the development of these skills in digital environments is not automatic. Instead, successful integration heavily depends on intentional pedagogical design and explicit teacher support. Significance: These findings underscore that technology alone is insufficient for language acquisition; rather, the field of English language education must prioritize cultivating self-regulated, strategic digital learners. This framework provides educators and curriculum designers with an actionable approach to foster learners’ independence and enhance linguistic proficiency in an evolving digital landscape.

Keywords: Metacognition, Learning Strategies, EFL, Technology-Enhanced Learning, Self-Regulated Learning  

UNDERSTANDING EFL TEACHERS’ ASSESSMENT BELIEFS IN INDONESIAN HIGHER EDUCATION: PREFERENCES, GRADING DECISIONS, AND QUALITY ASSURANCE 

Abstract ID: 259

Festif Rudolf Hoinbala

Artha Wacana Christian University 

Corresponding email: festifrudolf@gmail.com 

Teachers’ personal beliefs play a crucial role in shaping classroom-based assessment practices, yet empirical evidence from higher education EFL contexts in Indonesia remains limited. This study investigates Indonesian tertiary-level EFL teachers’ personal beliefs about language assessment, focusing on three dimensions: preferred assessment types and methods, the influence of student characteristics on grading decisions, and beliefs about assessment quality procedures. Using a quantitative survey design, data were collected from 197 EFL teachers across public and private universities in Indonesia through a structured questionnaire adapted from established assessment belief instruments. Descriptive statistical analyses reveal a strong endorsement of communicative and formative-oriented assessment practices, particularly oral presentations and feedback-rich assessment. Teachers also reported that student-related factors such as effort, general ability, and attitude frequently influence grading decisions, whereas demographic characteristics such as gender and age were perceived as largely irrelevant. Furthermore, respondents expressed near-unanimous agreement regarding core assessment quality principles, including validity, reliability, collaborative review, and transparent documentation. These findings suggest that Indonesian EFL teachers’ assessment beliefs are broadly aligned with contemporary learning-oriented assessment paradigms, while also highlighting ongoing tensions between holistic professional judgment and formal assessment standards. The study contributes to research on teacher cognition and language assessment literacy by providing context-sensitive evidence from higher education and underscores the importance of sustained professional development to strengthen belief–practice coherence in EFL assessment.

Keywords: EFL teachers; assessment beliefs; language assessment literacy; higher education; formative assessment 

GENDER AND ACADEMIC-LEVEL DIFFERENCES IN AI USE FOR ACADEMIC WRITING IN INDONESIAN HIGHER EDUCATION 

Abstract ID: 284

Ariyanti, Sukaena, Agnes Veronika

Universitas Widya Gama Mahakam Samarinda

Corresponding email: ariyanti@uwgm.ac.id

The rapid spread of AI chatbots has begun to reshape how university students plan, draft, revise, and evaluate academic writing. Although recent studies have examined students’ perceptions of generative AI in higher education, less is known about how students in regional Indonesian universities understand and use these tools, particularly when viewed across gender, age, and academic level. This study investigated the perceptions and reported use of AI chatbots among 267 students from four universities in East Kalimantan. A mixed-method survey design was used. Quantitative responses were analyzed through descriptive statistics and Spearman’s rho correlation, while open-ended responses were coded to identify the AI tools students reported using for academic writing. The results show that AI use was widespread, with 94.76% of respondents indicating that they had used AI in academic writing. Students generally held positive views of AI chatbots, especially for improving writing effectiveness, language quality, and the revision process. At the same time, their responses revealed continuing ethical and pedagogical uncertainty, particularly around learning, authorship, and the boundary between assistance and cheating. Correlation analysis showed weak relationships between students’ perceptions and gender, age, or semester, suggesting that AI chatbot use was broadly shared across demographic groups. The study offers regional evidence from Indonesian higher education and underscores the need for clearer AI literacy instruction, responsible-use guidelines, and process-based academic writing pedagogy.

Keywords: Academic rapid, AI Chatbots, Higher education