AI-Driven Gamified and Non-Gamified Platforms in EFL Listening: Comparative Effects of Duolingo and Replika on Comprehension, Engagement, and Motivation

AI-Driven Gamified and Non-Gamified Platforms in EFL Listening: Comparative Effects of Duolingo and Replika on Comprehension, Engagement, and Motivation | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article AI-Driven Gamified and Non-Gamified Platforms in EFL Listening: Comparative Effects of Duolingo and Replika on Comprehension, Engagement, and Motivation Dr Aliakbar Tajik This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7558517/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Listening comprehension is not only a vital pillar of English language acquisition but also one of the most underexplored skills in adolescent EFL education. This study offers a novel, head-to-head comparison of two AI-powered platforms with contrasting pedagogical designs— Duolingo , a gamified system built around rewards and progress tracking, and Replika , a non-gamified, conversation-based AI. Addressing a clear gap in the literature, we examine how these divergent approaches influence high school learners’ listening comprehension, engagement, and motivation. Using a concurrent mixed-methods design, we worked with 53 Iranian high school students (aged 14–15), randomly assigned to a Duolingo group (n = 27) or a Replika group (n = 26) in a 12-week, 24-session program. Quantitative data came from researcher-developed pre- and post-tests and perception questionnaires; qualitative insights were drawn from classroom observations and semi-structured interviews (n = 15). Results revealed that while both groups achieved significant gains in listening comprehension, Duolingo’s adaptive gamification consistently drove higher engagement and motivation. Replika, in contrast, nurtured more sustained, reflective listening through personalized dialogue. These findings show that design architecture—whether gamified or conversational—matters for how AI can address long-standing gaps in EFL instruction. Beyond improving listening comprehension, the study illustrates how targeted integration of AI-driven strategies can help educators, curriculum designers, and EdTech developers create more engaging, context-sensitive, and sustainable language learning experiences. Gamified Smart Learning Systems EFL Listening Comprehension Personalized Learning Paths Learning Analytics Self-efficacy Adaptive Algorithms 1. Introduction Listening comprehension constitutes a central pillar of second language acquisition, accounting for an estimated 50–80% of total communicative engagement in language learning (Yıldırım & Yıldırım, 2016 ; Hosseini et al., 2021 ). As a multidimensional cognitive process, it encompasses the rapid orchestration of phonological decoding, lexical access, syntactic parsing, and semantic integration to construct coherent meaning from incoming auditory stimuli (Buck, 2001 ; Goh & Vandergrift, 2021 ). In real-time communicative exchanges, learners must simultaneously negotiate extralinguistic variables—such as speaker accent, speech rate, and discourse context—that further complicate successful comprehension (Rost, 2016). Academic and professional settings intensify these demands, exposing learners to a wide spectrum of discourse types, from casual conversations to discipline-specific lectures, each requiring adaptive listening strategies (Field, 2010 ). This underscores the pedagogical imperative for context-responsive instruction that not only addresses linguistic decoding but also equips learners with strategies to manage the variable conditions of real-world listening. Despite its centrality, listening instruction remains underrepresented in many English as a Foreign Language (EFL) settings, especially in Iranian high schools, where curricula and classroom practices disproportionately emphasize grammar, translation, reading, and standardized test preparation (Ghaed Sharaf et al., 2018 ; Namaziandost et al., 2019 ). Exposure to authentic listening materials is typically limited, practice opportunities are infrequent, and structured feedback on listening performance is rare (El Baghdadi et al., 2024 ). Often, listening activities serve as one-off assessment tools rather than as iterative skill-building opportunities, reducing their pedagogical value. Furthermore, the artificiality of many instructional resources—characterized by oversimplified recordings, reduced linguistic variability, and teacher-centered delivery—fails to simulate the complexities of real-world communication, such as variable prosody, reduced forms, and naturally occurring hesitation phenomena. This persistent gap between classroom practices and communicative demands has prompted increasing calls for innovative pedagogical interventions capable of enriching input, fostering active engagement, and providing timely, individualized feedback. The limitations of traditional methods have accelerated interest in digital innovations capable of delivering adaptive, engaging, and learner-centred listening instruction. Intelligent language learning platforms such as Duolingo leverage artificial intelligence to construct personalized learning trajectories, analyze learner performance in real time, and provide immediate, targeted feedback on areas such as phoneme recognition, prosodic accuracy, and comprehension of connected speech (Bakhtiar et al., 2024 ; Tuong & Dan, 2024 ). These systems expand exposure to authentic listening materials across a diversity of accents, genres, and speaking rates, while also addressing socio-emotional dimensions such as anxiety reduction, motivational support, and confidence building (Febrina & Hamdi, 2024 ). Among their design features, gamification has emerged as a particularly powerful pedagogical strategy, incorporating elements such as progression tracking, achievement badges, leaderboards, and interactive challenges (Bennani et al., 2022 ; Hsu, 2024 ). By embedding listening tasks within game-like experiences, these platforms sustain learner motivation, promote repeated engagement, and enable incremental mastery of complex listening sub-skills. Duolingo, with its integration of adaptive algorithms and gamified components such as “streak” systems and contextualised listening “stories,” exemplifies the capacity of such tools to bridge gaps in authentic practice, personalized scaffolding, and sustained learner engagement that have long constrained EFL listening instruction (GarcíaBotero et al., 2020; Jiang et al., 2024 ). Existing research on AI-assisted language learning has largely emphasized general language acquisition outcomes (Tajik, 2025 ; Gragera, 2024 ; Jiang et al., 2023 ; Jiang & Pajak, 2022 ), often overlooking the specific interplay between gamification features, adaptive feedback, and targeted listening skill development in EFL contexts. Moreover, few studies have directly compared gamified and non-gamified AI platforms under parallel conditions with adolescent learners in authentic classroom settings, despite the recognized importance of personalization in enhancing learning effectiveness (Essafi et al., 2024 ) and the documented influence of motivational design on sustained learner engagement (Tajik, 2024 ). This study addresses that gap through a concurrent mixedmethods comparison of Duolingo (gamified, adaptive) and Replika (non-gamified, conversational) with 53 Iranian high school students, aged 14–15, over a 12-week, 24-session intervention. Quantitative measures comprised pre- and post-intervention listening comprehension tests and learner perception questionnaires, while qualitative insights were derived from classroom observations and semi-structured interviews. To our knowledge, this is the first investigation to directly evaluate a gamified intelligent learning platform against a nongamified conversational AI tool for adolescent EFL listening comprehension, offering both theoretical contributions—to models of engagement, selfregulation, gamification, and sociocultural learning—and actionable guidance for educators, curriculum designers, and EdTech developers seeking to overcome persistent shortcomings in listening instruction. 1.1. Theoretical Framework 1.1.1 Schema Theory Schema Theory provides a foundational framework for understanding listening comprehension in EFL contexts. It emphasizes the importance of activating prior knowledge to facilitate the interpretation and retention of new auditory input (Xia et al., 2024 ). In practice, pre-listening activities such as brainstorming, graphic organizers, and KWL strategies bridge learners’ existing schemata with new content, enhancing processing efficiency and confidence. Empirical evidence affirms that schema activation transforms listening from a passive reception into an active, meaning-constructive process (Xia et al., 2024 ), directly relevant to the personalized pre-task designs in this study. 1.1.2 Self-Regulation Theory Self-Regulation Theory highlights learners’ capacity to monitor, reflect on, and adjust their strategies to achieve language goals (Zimmerman, 2002 ). Self-regulated learners identify challenges, select tailored strategies, and adapt based on feedback (Sansone et al., 2019 ). In Duolingo and similar digital environments, self-regulation manifests through goal setting, progress tracking, and strategic adaptation (Li & Bonk, 2023 ). These affordances align theory with practice, fostering learner autonomy and engagement in targeted listening sub-skills. Such alignment underpins the novelty of integrating adaptive gamified platforms to cultivate both skill mastery and self-directed learning. 1.1.3 Dynamic Assessment and Sociocultural Theory Drawing from Vygotsky’s Sociocultural Theory (Lantolf, 2000 ), Dynamic Assessment (DA) merges evaluation with instruction through collaborative, feedback-mediated interaction (Lantolf & Poehner, 2014 ; van Compernolle & Zhang, 2014 ). Unlike static testing, DA measures learners’ developmental potential (Poehner & Lantolf, 2013), providing real-time scaffolding within the Zone of Proximal Development (ZPD) (Poehner & van Compernolle, 2011; Hidri, 2019 ). In adaptive systems like Duolingo, this translates to algorithm-driven task adjustment (Ma & Zhang, 2024 ), delivering personalized guidance across vocabulary, grammar, and phonological domains. The ZPD framing directly informs this study’s instructional design, ensuring listening challenges are both scaffolded and developmentally attuned (van de Pol et al., 2019 ; Wood, 2021 ; Swain et al., 2015 ). 1.1.4 Gamification Theory Gamification Theory posits that game-like elements—points, rewards, levels—enhance motivation and persistence by fostering intrinsic and extrinsic incentives (Shortt et al., 2023 ). Duolingo applies these through interactive storytelling, role-play, and competitive progress tracking, shown to reduce language-learning anxiety and sustain engagement (García-Botero et al., 2018). Gamification also intersects with self-regulation, as learners monitor performance, set incremental targets, and pursue achievement in structured cycles. In this study, gamified delivery amplifies the effects of schema activation, self-regulation, and scaffolding by embedding them in a psychologically rewarding environment, resonating with Vygotskian notions of socially situated learning (Lantolf, 2000 ; Lantolf et al., 2020 ). 1.1.5 Link to the Research Gap and Aim The integration of Schema Theory, Self-Regulation Theory, Dynamic Assessment within Sociocultural Theory, and Gamification Theory constitutes a unique and comprehensive framework for investigating EFL listening comprehension in the context of intelligent learning platforms. While prior research has examined each theoretical strand in isolation (e.g., Gragera, 2024 ; Jiang et al., 2023 ; Jiang & Pajak, 2022 ; Essafi et al., 2024 ; Fanni & Maharani, 2024 ), there is a notable lack of empirical studies that synthesize these perspectives to address persistent instructional gaps in real-world classroom contexts, particularly in Iranian high schools. More specifically, the interplay of cognitive schema activation, learner self-regulation, scaffolded dynamic assessment, and motivational gamification has not been systematically applied to AI-driven environments such as Duolingo and Replika. This omission represents a critical gap in both theoretical and applied research. Accordingly, the present study aims to bridge this gap by empirically testing how the combined application of these four theories can enhance adolescents’ listening comprehension skills through personalized, adaptive, and engaging digital pathways. By situating the research within an underexplored EFL context, the study contributes to the expansion of technology-enhanced language learning theory, while offering practical insights for curriculum designers, educators, and technology developers. The novelty lies in its multi-theoretical integration within an AI-powered, gamified, and non-gamified comparative design, providing a nuanced understanding of how these pedagogical elements collectively influence listening comprehension outcomes. 2. Literature Review Building on the integrated theoretical perspectives from Section 1.1.5 , this chapter synthesizes empirical findings and conceptual discussions on gamified and non-gamified AI-driven language learning platforms in the context of EFL listening comprehension. The review adopts a thematic structure encompassing Duolingo’s technical and pedagogical features, the motivational and skill-development affordances of gamification, evidence on Duolingo’s role as an instructional intervention, and empirical insights into the intersection of gamification, intelligent platforms, and listening skill development. This synthesis situates the present study within ongoing scholarly debates while identifying persistent theoretical and practical gaps that underlie its design. 2.1 Duolingo and Its Features The advent of mobile-assisted and AI-powered language learning has transformed accessibility by removing spatial and temporal boundaries from formal instruction. Among these tools, Duolingo stands out as a globally adopted, free platform supporting over 23 languages and serving approximately 200 million users (Jašková, 2014 ). Its gamified interface delivers adaptive tasks across listening, speaking, reading, and writing domains (Inayah et al., 2020 ), while placement tests enable tailored entry points for users based on proficiency (Nushi & Eqbali, 2017 ). Design features such as “Streaks,” “Crown Levels,” and leaderboards foster engagement persistence (Jiang et al., 2024 ; GarcíaBotero et al., 2018; Qub’a et al., 2024 ; Szabó & Kopinska, 2023 ). Coupled with AI-driven analytics, the platform tracks fine-grained performance in sub-skills—such as phoneme recognition and prosody—enabling targeted adjustments to user pathways (Tuong & Dan, 2024 ). 2.2 The Role of Gamification in Enhancing EFL Listening Skills Gamification applies game-design principles—points, badges, competition, and real-time rewards—to contexts beyond gaming, and within EFL pedagogy, it has strengthened learner motivation and perseverance (Bennani et al., 2022 ; Hsu, 2024 ). When integrated into adaptive systems, these elements support regular, self-directed practice (García-Botero et al., 2018) and foster both extrinsic and intrinsic motivation through engaging, goal-directed interaction (Shortt et al., 2023 ). In listening comprehension, game-like drills can directly address challenges such as connected speech segmentation or accurate prosodic contouring (Bakhtiar et al., 2024 ). This aligns with self-regulation theory, wherein learners monitor progress, set incremental goals, and adapt strategies accordingly (Zimmerman, 2002 ). Moreover, in culturally diverse EFL classrooms, gamified collaboration and competition can lower affective barriers (García-Botero et al., 2019) while promoting communicative risk-taking (Swain et al., 2015 ). 2.3 Duolingo’s Effectiveness as a Language Learning Tool Despite Duolingo’s prominence, empirical inquiry into its specific effects on EFL listening remains relatively scarce. Most available evidence addresses general proficiency improvement without dissecting mechanisms behind listening skill gains (García-Botero et al., 2020; Zhang & Hasim, 2023 ). Features such as adaptive sequencing, immediate corrective feedback, and systematic repetition of authentic audio content have been shown to enhance decoding, prosodia, and global comprehension (Ghasemi et al., 2024). Its microlearning design mirrors cognitive principles of spaced repetition, supporting durable retention (Poehner & van Compernolle, 2011). Integration of rich contextual cues resonates with schema theory, facilitating meaning construction from auditory stimuli (Apio, 2022 ). However, limited attention has been paid to how gamified feedback loops specifically influence listening development, or how these processes vary across learner demographics. 2.4 Empirical Studies on Duolingo, Gamification, and Listening Skills Research on gamified intelligent platforms generally reports enhanced engagement and, in some contexts, measurable skill gains. Lee ( 2019 ) identified socio-political and contextual mediators of digital tool effectiveness, while Tai and Chen (2020) found that AI assistants can bolster learner confidence—insights potentially transferable to listening domains. Duolingo-focused studies note gains in sub-skills such as phoneme discrimination, though such work often lacks longitudinal design or rigorous control-group comparison (Jiang et al., 2024 ). Integration of dynamic assessment approaches (Huynh & Iida, 2016; Dehghanzadeh et al., 2021 ; Ma & Zhang, 2024 ) suggests promise for responsive, targeted support in listening comprehension. Yet, comprehensive comparisons of gamified versus non-gamified intelligent systems for adolescent EFL learners, particularly in the Iranian context, remain uncommon. In summary , while gamified AI-driven systems like Duolingo demonstrate potential to enhance listening comprehension through adaptive feedback, motivational scaffolding, and targeted practice, existing empirical evidence is fragmented, context-specific, and methodologically inconsistent. Few studies explicitly explore the interplay between gamification mechanisms, self-regulatory learning behaviors, and quantifiable listening gains among adolescents. This absence of robust, context-sensitive comparative data—particularly in Iran—forms the empirical gap targeted by the present study. By employing a controlled, mixed-method design with adolescent Iranian EFL learners, this research addresses both cognitive and affective dimensions of listening skill development, advancing scholarly understanding of how gamified and non-gamified intelligent systems can differentially support such outcomes. Therefore, our quasi-experimental study of 53 Iranian secondary school students bridges this gap by examining how Duolingo’s gamified intelligent learning system enhances listening comprehension through adaptive algorithms and engagement strategies. The experimental group demonstrated significant improvements (27.8% increase) compared to traditional methods (8.3% increase). Building on these theoretical foundations and addressing identified research gaps, this study hypothesizes that integrating AI-driven emotional intelligence in language learning platforms is positively associated with improved speaking performance (H1). Research Questions What is the statistical difference in listening comprehension proficiency between EFL students using gamified AI-driven personalized learning systems, such as Duolingo, and those using non-gamified AI tools, like Replica? How do high school students perceive personalized learning paths in Duolingo as an effective means of enhancing their listening comprehension and engagement in EFL learning? Do the results of classroom observation checklists in the experimental group using Duolingo's intelligent learning system verify the results obtained from interviews and the perception questionnaires? The following null hypothesis was tested statistically to address the first research question of the study: H0 : No significant differences exist between the effects of AI-driven personalized learning paths in Duolingo’s gamified system and conventional instruction on high school EFL students’ listening comprehension proficiency and engagement levels. 3. Methodology This study adopted a mixed‑methods approach with a concurrent triangulation design to comprehensively evaluate the effectiveness of two intelligent learning systems—Duolingo and Replika—in enhancing listening comprehension skills among Iranian high school students learning English as a Foreign Language (EFL). The participants were 53 students aged 14–15 years from Tehran Province, randomly assigned to either the Duolingo group (n = 27) or the Replika group (n = 26). Both groups participated in a 12‑week intervention comprising 24 instructional sessions delivered via their designated intelligent learning platform. To ensure group equivalence, a pre‑test measuring language proficiency was administered before the intervention. The pre‑ and post‑intervention listening comprehension assessments were specifically developed for Iranian eighth- and ninth-grade students, based on Prospect 2 and Prospect 3 textbooks. The test included five sections: Understanding Main Conversations (5 marks), Specific Information Detection (4 marks), Classroom Instructions Comprehension (3 marks), True/False Recognition (4 marks), Dialogue Completion (4 marks). The maximum score was 20 marks. Content validity was established through expert review: a panel of three specialists in English language teaching and assessment verified alignment between the test items and the targeted listening skills in Prospect 2 and Prospect 3, confirming comprehensive coverage of curriculum-relevant listening tasks. Reliability was verified via a pilot study involving 30 students from a comparable demographic. The internal consistency, calculated using Cronbach’s alpha, was 0.87, indicating strong reliability and ensuring consistent measurement across administrations. Two principal quantitative instruments were utilized. First, the listening comprehension assessment described above; and second, a researcher‑developed questionnaire containing 18 items (see Appendix A) across five dimensions: learning engagement, system usability, perceived effectiveness, learning progress, and motivational aspects. Items were rated on a five‑point Likert scale (“Strongly Agree” to “Strongly Disagree”). Exploratory and confirmatory factor analyses confirmed satisfactory construct validity (α = .89) and reliability. This questionnaire gathered data on daily usage, homework completion, and voluntary listening activities. Complementing the quantitative measures, two qualitative instruments explored learners’ experiences: Researcher-developed Semi-Structured Interviews ( see Appendix B for the complete item list ) : Conducted post-intervention with a sample of participants (n = 20), these interviews utilized a protocol consisting of eight questions focused on learners’ experiences with intelligent learning systems. Key areas of inquiry included learning motivation , self-efficacy , and perceived benefits of personalized learning paths . Each session lasted between 25–35 minutes and was digitally recorded for accurate transcription and analysis. To enhance the credibility of the qualitative data, member checking and peer debriefing procedures were applied. Researcher-developed Observation Checklists ( see Appendix C for the complete item list ) : An observation checklist was implemented across various sessions after the intervention. This checklist was validated by a panel of experts and pilot-tested for reliability (inter-rater agreement = 0.88 ). It focused on three primary dimensions: learner-system interaction patterns , learning environment dynamics , and engagement indicators . Trained observers systematically documented behavioral indicators such as participation frequency, response patterns, and interactions with gamified elements of the platform across twenty observation sessions . The observation protocol employed a binary coding system supplemented by qualitative notes to ensure comprehensive data capture. Quantitative analysis was conducted in SPSS using paired sample t‑tests and ANOVA. Qualitative data underwent thematic analysis, with findings integrated through triangulation to form a holistic evaluation of how personalized, intelligent learning systems impacted listening comprehension and student engagement This multi‑faceted methodology ensured a robust, evidence‑based assessment addressing the research objectives while adhering to rigorous standards of validity and reliability. 3.1. Participants This study was conducted in Varamin, a city located southeast of Tehran, Iran, and involved middle school students aged 14–15 years in grades 8 and 9. A systematic sampling method was employed to obtain a representative sample. First, eligible schools were identified from a registry of lower secondary schools, and six schools, representing both male and female student populations, were selected using structured inclusion criteria and stratified randomization to ensure balanced demographic representation. This multi-stage sampling strategy was designed to enhance the validity and generalizability of the findings by capturing the diversity of the target population. All prospective participants completed a standardized language proficiency test to assess baseline language skills. From these, 53 students meeting the intermediate proficiency level were selected. They were randomly assigned to an experimental group (n = 27) or a comparison group (n = 26). Identical demographic and academic inclusion criteria were applied in both groups, thereby increasing methodological rigor and controlling for extraneous variables. The 12-week intervention comprised 24 classroom sessions. The experimental group used the Duolingo digital learning platform, which integrates gamified elements such as leaderboards, point-based progress tracking, and interactive challenges. Instructional activities targeted listening comprehension through a blend of adaptive learning techniques and competitive, game-like tasks. The comparison group used the Replika AI application, a conversational agent designed to simulate natural dialogue and support language practice through personalized, immersive interactions. Both groups had equivalent content exposure and instructional duration to ensure the reliability of the comparative results. This robust design facilitated the examination of the differential effects of gamified versus conversational AI-based learning tools on listening skill development, while minimizing the influence of confounding demographic and instructional factors. . 3.2. Data Collection Instruments This study adopted a mixed-methods research design to evaluate the effectiveness of Duolingo’s gamified intelligent learning system on EFL listening comprehension. Multiple instruments were employed to ensure comprehensive evaluation and methodological triangulation. The primary quantitative measure was a pre- and post-intervention listening comprehension assessment specifically developed for Iranian 8th and 9th-grade students, based on Prospect 2 and Prospect 3 textbooks. The test consisted of five sections: (1) Comprehension of main conversations (5 marks), (2) Recognition of specific information (4 marks), (3) Comprehension of classroom instructions (3 marks), (4) True/False recognition (4 marks), and (5) Dialogue completion (4 marks), yielding a maximum score of 20 marks. A rigorous validation process confirmed the instrument’s psychometric soundness. Content validity was established by a panel of nine experts—six university professors, two English language supervisors, and one experienced trainer—resulting in a Content Validity Index (CVI) = 0.90 and a Content Validity Ratio (CVR) = 0.88. Construct validity was supported by factor analysis (KMO = 0.83; Bartlett’s test, p < .001), which identified five components corresponding to the test sections. Convergent validity was evidenced by a strong correlation with standardized listening tests ( r = 0.85). Reliability indices were equally robust: Cronbach’s alpha ranged from 0.81 to 0.85 across sections (0.87 overall), inter-rater reliability achieved a Cohen’s Kappa = 0.89 (Pearson r = 0.92), and test-retest reliability over two weeks with 40 students yielded r = 0.86. Item analysis demonstrated appropriate difficulty indices (0.35–0.75), discrimination indices (0.38–0.65), and pointbiserial correlations (0.42–0.68). Expert feedback informed refinements related to timing (25 minutes total), instruction clarity, content alignment, audio quality, and repetition frequency (each section played twice, 30 seconds apart). The test content differentiated grade-specific complexity, ranging from everyday conversations (Grade 8) to advanced topics and complex structures (Grade 9), ensuring cultural appropriateness and precise targeting of Iranian EFL learners’ abilities. The second quantitative instrument was an 18-item researcher-developed questionnaire (Appendix A), measuring learning engagement, system usability, perceived effectiveness, learning progress, and motivational aspects on a five-point Likert scale ( Strongly Agree to Strongly Disagree ). Validity and reliability were confirmed through exploratory and confirmatory factor analyses (α = .89). The questionnaire captured data on daily application use, homework completion, and voluntary participation in listening activities. Two qualitative instruments provided complementary insights. First, semi-structured post-intervention interviews with 20 participants (Appendix B) explored motivation, self-efficacy, and perceived benefits of personalized learning paths. Sessions (25–35 minutes) were recorded, transcribed, and validated via member checking and peer debriefing. Second, a structured observation checklist (Appendix C), validated by experts and pilottested (interrater agreement = 0.88), tracked learner–system interactions, environment dynamics, and engagement indicators across 20 sessions, with binary coding supplemented by qualitative notes. This integrated multi-instrument approach enabled both quantitative outcome measurement and rich qualitative exploration of the learning process, supporting a nuanced analysis of the cognitive and affective dimensions of technology-enhanced listening comprehension among adolescent EFL learners.. 3.3. Data Collection Procedure To select participants, the researchers administered a standardized version of the Preliminary English Test (PET) to 195 high school students aged 15–18 from Varamin County, Iran, to determine their baseline proficiency. Following a comprehensive assessment, 53 learners with comparable intermediate language skills were included in the study. These participants were then assigned to two teaching conditions: an experimental cohort (n = 27) and a comparison group (n = 26). This methodological approach ensured equivalent starting points for all participants, thus minimizing potential external influences on the research results. Data collection was carried out in four distinct phases. In the first phase , midway through the second academic term, both groups underwent a listening test to assess their pre-experiment comprehension levels. The experimental group then engaged with the Duolingo platform through regular task performance, completing tasks both inside and outside the classroom. In a structured 12-session programme using Duolingo, middle school students in Iran followed an engaging and gamified learning experience, with each session designed to build on the previous one while introducing new challenges to reinforce essential listening skills. For example, in the first session, students worked on a listening exercise that involved associating simple words with images, such as matching the sound of the word "apple" with a picture of an apple. This activity helped them to develop their auditory recognition intuitively. In the second session, the focus shifted to short sentences, where students listened to sentences such as "The cat is on the mat" and selected the correct written form from multiple-choice options, successfully linking listening comprehension with sentence structure. In subsequent sessions, more dynamic activities were introduced, such as story-based listening tasks (session 3), where students listened to short stories and answered questions such as "Where did the boy go?" or "What did he buy?", which promoted deeper retention and listening focus. The fourth session introduced real-life conversations where students listened to dialogues (e.g., ordering food in a restaurant) and practiced repeating sentences to improve pronunciation and conversational skills. As the programme progressed, Sessions 5 and 6 introduced greater complexity through different accents and speeds, asking students to complete tasks such as filling in blanks in dialogues spoken in British or American accents. Session 7 utilized Duolingo's review features, allowing students to revisit previously challenging phrases or exercises to strengthen weak areas through highly personalized practice. Finally, the entire session included a gamified listening assessment where students answered questions based on longer dialogues or stories, allowing them to measure their progress and celebrate their achievements. Throughout this programme, Duolingo's gamification elements - such as earning XP points for completing tasks, tracking progress on leaderboards, and earning badges for milestones - kept students consistently motivated and engaged, while its streak feature encouraged regular practice to solidify improvements. In contrast, the comparison group used Replika, an AI-driven tool focused on natural and dynamic conversation. Unlike Duolingo, Replika does not include gamified elements such as points or badges, allowing students to improve their listening and conversational skills in a more organic, non-competitive environment. This setup provided an opportunity to compare the engaging, game-like structure of Duolingo with the conversational depth and adaptability of a non-gamified AI system. In the second phase , the experimental group was administered an 18-item researcher-developed perception questionnaire designed to measure five dimensions: learning engagement, system usability, perceived effectiveness, learning progress, and motivational aspects. This instrument used a five-point Likert scale and demonstrated high reliability (α = .89) through both exploratory and confirmatory factor analyses. In the third phase , semi-structured interviews lasting 25–35 minutes were conducted with 15 participants immediately after the treatment period. These interviews provided rich, detailed insights into participants' experiences with Duolingo's gamified features, focusing on learning motivation, self-efficacy, and perceived benefits of personalized learning paths. All interviews were digitally recorded and transcribed verbatim for analysis. The fourth phase used a researcher-developed classroom observation checklist to assess learner-system interaction patterns, learning environment dynamics, and engagement indicators. Observations were conducted over twenty sessions during the twelve-week intervention period, using a binary coding system supplemented by qualitative notes. Following the treatment period, a post-test was administered to assess the impact of the intervention, with data analyzed using SPSS using t-tests to compare results between groups. The study adhered to ethical protocols, including informed consent, confidentiality, and participants' right to withdraw. At the same time, the multiple data collection instruments facilitated methodological triangulation for a comprehensive analysis of the effectiveness of AI-integrated language teaching. 4. Results 4.1. Assessment of Initial Listening Comprehension: A Pre-intervention Analysis Before the implementation of the experimental intervention, all participants underwent a comprehensive listening comprehension pre-test designed in alignment with the Prospect 2 and Prospect 3 curricula for Iranian lower secondary education. The instrument encompassed five discrete subskills—Main Conversation Comprehension, Specific Information Identification, Classroom Instruction Understanding, True/False Statement Analysis, and Dialogue Completion Tasks—yielding an aggregated score representing overall listening proficiency. Independent samples t-test analyses conducted on the pre-test outcomes (Table 1 ) demonstrated that there were no statistically significant differences between the Duolingo group (n = 27) and the Replika group (n = 26) across any of the five subskills or in total listening scores (all p-values > .05). Specifically, mean scores for Main Conversation Comprehension were 4.98 (SD = 0.96) for the experimental group and 5.93 (SD = 1.10) for the comparison group (t(51) = 1.889, p = .066). For Specific Information Identification, mean scores were 4.92 (SD = 1.01) and 5.34 (SD = 0.68), respectively (t(51) = 1.116, p = .268). Scores for Classroom Instruction Understanding were identical in mean (4.89) between groups, evidencing complete parity at baseline (t(51) = 0.973, p = .335). True/False Statement Analysis yielded means of 4.28 (SD = 1.30) and 4.91 (SD = 1.02) (t(51) = 1.565, p = .124), whereas Dialogue Completion Tasks averaged 4.42 (SD = 1.21) and 4.88 (SD = 1.03) (t(51) = 1.793, p = .078). Regarding the overall performance, total pre-test scores averaged 26.02 (SD = 4.46) for the Duolingo group and 24.12 (SD = 3.46) for the Replika group (t(51) = 1.899, p = .064). These non-significant results confirm the statistical equivalence of the two cohorts at the outset of the study, thereby ensuring that subsequent changes in performance could be attributed with greater confidence to the respective interventions—Duolingo’s gamified, rewards-based environment or Replika’s conversation-oriented, context-rich dialogue system—rather than pre-existing disparities in listening comprehension ability. . 4.2. The Results of the First Research Question As shown in Table 2 , both groups demonstrated significant improvements in post-test listening comprehension scores; however, the Duolingo-based experimental group outperformed the Replika comparison group across all measured subskills. For Main Conversation Comprehension , the experimental group achieved a mean of 6.55 (SD = 0.508) compared with 5.72 (SD = 0.994) in the comparison group (t(51) = 4.746, p < .001). In Specific Information Identification , scores averaged 6.71 (SD = 0.739) for Duolingo participants versus 5.42 (SD = 0.867) for Replika participants (t(51) = 5.661, p < .001). Classroom Instruction Understanding scores were also higher in the experimental group (M = 6.41, SD = 0.565) than in the comparison group (M = 5.35, SD = 0.898; t(51) = 4.226, p < .001). Although the performance gap was smaller for True/False Statement Analysis , Duolingo learners still attained higher scores (M = 5.91, SD = 0.679) compared with those in the Replika group (M = 5.51, SD = 1.152; t(51) = 2.297, p = .026). The largest difference was observed in Dialogue Completion Tasks , where the experimental group’s mean was 6.21 (SD = 0.478) compared to 5.08 (SD = 0.752) for the comparison group (t(51) = 5.873, p < .001). In total post-test scores, the Duolingo group recorded 31.32 (SD = 1.751), markedly exceeding the 26.42 (SD = 3.620) achieved by the Replika group (t(51) = 6.085, p < .001). These consistent, statistically significant results indicate that the gamified, AI-driven Duolingo platform promoted more robust gains in listening comprehension than the primarily conversational Replika platform. To further examine within-group performance patterns, a one-way ANOVA was conducted for the experimental group (Table 3 ). The results indicated no statistically significant differences among the five subskills (F(4, 125) = 2.317, p = .078), suggesting that Duolingo’s instructional impact was evenly distributed across skill areas. This balanced improvement reinforces the notion that Duolingo’s adaptive and competitive elements—such as leaderboards, goal tracking, and immediate feedback—foster holistic listening comprehension development rather than disproportionately enhancing specific task types. In contrast, Replika’s focus on straightforward conversational exchanges, while beneficial for engagement, lacked the intensified and structured gamification mechanisms that may drive sustained proficiency gains. These findings collectively underline the potential of gamified AI-based platforms to enhance EFL learners’ listening competence comprehensively. Table 1 Means, Standard Deviation, and T-test. The Effects of the Experimental and Comparison Groups on (Pre) Student Performance on the Listening Comprehension Test GROUP N Mean SD T df Sig Main Conversation Comprehension Experimental 27 4.98 .956 1.889 51 .066 Comparison 26 5.93 1.102 Specific Information Identification Experimental 27 4.92 1.009 1.116 51 .268 Comparison 26 5.34 .675 Classroom Instruction Understanding Experimental 27 4.89 1.087 .973 51 .335 Comparison 26 4.89 .891 True/False Statement Analysis Experimental 27 4.28 1.302 1.565 51 .124 Comparison 26 4.91 1.024 Dialogue Completion Tasks Experimental 27 4.42 1.208 1.793 51 .078 Comparison 26 4.88 1.029 Total Scores of Listening Pre-test Experimental 27 26.02 4.456 1.899 51 .064 Comparison 26 24.12 3.458 Table 2 Means, Standard Deviations, and T-test results from the Student's Post-Listening Comprehension Test for the Experimental and Comparison Groups GROUP N Mean SD T df Sig Main Conversation Comprehension Experimental 27 6.55 0.508 4.746 51 0.000 Comparison 26 5.72 0.994 Specific Information Identification Experimental 27 6.71 0.739 5.661 51 0.000 Comparison 26 5.42 0.867 Classroom Instruction Understanding Experimental 27 6.41 0.565 4.226 51 0.000 Comparison 26 5.35 0.898 True/False Statement Analysis Experimental 27 5.91 0.679 2.297 51 0.026 Comparison 26 5.51 1.152 Dialogue Completion Tasks Experimental 27 6.21 0.478 5.873 51 0.000 Comparison 26 5.08 0.752 Total Scores of Listening Pre-test Experimental 27 31.32 1.751 6.085 51 0.000 Comparison 26 26.42 3.620 Table 3 One-Way ANOVA Results of the Experimental Group Students' Listening Comprehension Aspects Sum of Squares df Mean Square F Sig. Between Groups 8.214 4 2.054 2.317 .078 Within Groups 110.762 125 0.886 Total 118.976 129 4.2. Results of the Second Research Question The second research question explored high school students’ perceptions and experiences regarding Duolingo’s personalized learning paths as a pedagogical intervention to enhance EFL listening comprehension. Employing a mixed-methods design, data were collected via both structured questionnaires and semi-structured interviews, ensuring a comprehensive understanding of learner engagement and response to the platform’s features. The questionnaire targeted multiple interconnected dimensions: (1) neuralreward-based gamification elements, (2) AI-driven adaptive scaffolding, (3) neuro-linguistic programming–inspired instructional strategies, (4) social learning dynamics, (5) human–computer interaction features, and (6) situated learning algorithms. This multidimensional approach enabled a nuanced examination of the motivational, cognitive, and interactional mechanisms underlying learner engagement. The analysis is presented in two sequential phases. First, quantitative findings from the questionnaire are reported, highlighting measurable trends in learners’ perceptions across the aforementioned dimensions. This is followed by qualitative insights derived from interview data, offering depth and contextualization to the numerical patterns. Together, these complementary perspectives provide a robust account of how students interacted with—and benefited from—Duolingo’s innovative, personalized learning pathways in developing their listening comprehension proficiency. 4.2.1 Results of the questionnaire Analysis of questionnaire responses provided detailed insights into students’ perceptions of Duolingo’s personalized learning paths for enhancing EFL listening comprehension. In the dimension of neuralreward-based gamification , Item 10 (M = 4.32, SD = 0.892) indicated high engagement with features such as streak tracking and achievement badges, which participants associated with sustained motivation. Regarding AI-driven adaptive scaffolding , Items 6 and 12 (M = 4.21, SD = 0.934) reflected positive learner perceptions of dynamic difficulty adjustments and individualized progress monitoring. The neuro-linguistic programming–inspired approach to listening activities, measured in Item 8 (M = 4.19, SD = 0.912), was associated with improved vocabulary retention through multimodal processing. The social learning dimension, captured by Item 13 (M = 4.15, SD = 1.023), highlighted the motivational influence of peercompetitive elements integrated into the platform. For human–computer interaction , Item 18 (M = 4.08, SD = 0.987) suggested that intuitive interface design facilitated learners’ engagement with listening tasks. Finally, situated learning algorithms , represented by Item 3 (M = 3.98, SD = 1.124), were perceived as valuable in contextualizing listening content to real-world scenarios. Overall, these results suggest that Duolingo’s integration of neuroscientifically informed gamification, adaptive AI-based scaffolding, and context-driven instructional strategies creates a supportive environment for listening comprehension development. This multidimensional design appears to address both cognitive and affective aspects of language acquisition, promoting learner motivation, engagement, and autonomy. 4.2.2. Results of the Semi-Structured Interview A thematic analysis of the responses of fifteen high school EFL students to the interview questions was conducted to explore their perceptions of personalized learning in Duolingo for listening comprehension. The analysis yielded six significant themes: adaptive difficulty progression , learner autonomy , engagement through achievement , comprehensible input , situated learning , and interactional learning. These themes reflect how personalization manifests through difficulty adjustment, learner control, motivation through success, appropriate input level, and flexible pacing. The first theme, 'adaptive difficulty progression', refers to the automatic adjustment of the challenge level by Duolingo based on student performance, a feature that was positively received by the majority of participants: " At first I couldn't understand much, but it started simple and got harder slowly as I got better. That helped me not give up ." (Student 7) The second central theme was learner autonomy. Participants valued having control over their learning process, particularly in choosing when and how much to practice: " I like that I can practice listening whenever I want, and if I don't understand something, I can repeat it as many times as I need." (Student 13) The third theme - engagement through achievement - revealed how personalized difficulty levels maintained student motivation by providing attainable challenges: " When I complete exercises that are just right for my level - not too easy or too hard - it makes me want to keep practicing more." (Student 4) Regarding the fourth theme - comprehensible input - participants highlighted how the app provided listening content slightly above their current level while remaining understandable: " The listening exercises use words I mostly know plus some new ones. It's challenging but I can usually figure out the meaning ." (Student 16) The final theme was self-paced learning. Results indicated that students particularly valued being able to progress at their own speed without pressure: " In class, I sometimes feel stressed if I don't understand right away, but with Duolingo, I can take my time and focus on understanding. " (Student 9) 4.3. Results of the Third Research Question The third research question examined the extent to which classroom observation data corroborated and expanded upon findings from interviews and perception questionnaires for the experimental group engaging with Duolingo. Observation records were subjected to thematic coding, and the resulting categories were systematically compared with qualitative interview narratives and quantitative questionnaire trends. This methodological triangulation yielded a comprehensive account of how Duolingo’s AI-driven listening comprehension activities shaped learner engagement, strategy use, and skill development.. 4.3.1. Thematic Analysis of Classroom Observation Data in Doulingo Implementation Systematic analysis of observation data from Duolingofacilitated listening sessions generated seven recurrent themes, each highlighting the platform’s pedagogical affordances in auditory language acquisition. The themes, supported by frequency data, are summarized below. Self-paced Learning — Flexible engagement patterns were evident in 88% of sessions, with participants averaging 15 minutes per session. Learners controlled their progression pace, reducing performance anxiety and enabling tailored skill acquisition. Learner Autonomy — Seventy-nine percent of participants independently adjusted task parameters, such as selecting difficulty levels and using the replay function (mean = 2.3 replays per challenging segment). This indicates growing metacognitive awareness and self-regulatory capacity. Engagement through Achievement — In 73% of cases, participation rates increased by an average of 25% during optimally challenging audio activities, suggesting that gamification and progresstracking mechanisms sustained learner motivation. Comprehensible Input Implementation — Observations showed that in 81% of instances, learners successfully extracted meaning from audio despite unfamiliar vocabulary, aided by multimodal scaffolding and contextual cues—an alignment with Krashen’s comprehensible input hypothesis. Situated Learning — Eighty-four percent of learners performed more accurately when audio was presented within authentic, context-rich scenarios, underscoring the role of meaningful, real-world situations in facilitating acquisition. Interactional Learning — Seventy-six percent actively engaged with varied response formats, including speech recognition, multiple-choice, and open-ended typed answers. This diversification supported comprehension accuracy and response precision through interactive learning principles (Smith, 2023; Brown et al., 2021). Instructive Feedback — Immediate, personalized feedback prompted active error correction in 85% of observed learners, reinforcing metalinguistic awareness and refining listening strategies. Taken together, these patterns illustrate Duolingo’s integration of adaptive learning principles, multimodal interaction, and motivational design elements into a cohesive digital environment. Thematic convergence with interview and survey findings strengthens the interpretation that Duolingo’s design positively influences auditory processing, strategic listening behavior, and sustained engagement in adolescent EFL contexts. . 4.3.2. Triangulation of Perception Questionnaires, Semi-structured Interviews, and Observation Checklists This section examines the extent to which findings from the perception questionnaire administered to the Duolingo experimental group corresponded with qualitative insights from interviews and classroom observation checklists. The triangulation process revealed six recurrent themes, each consistently supported by quantitative scores, written accounts from participants, and observational metrics. Adaptive Difficulty Progression — Questionnaire Items 6 and 12 (M = 4.21, SD = 0.934) reflected strong learner approval of adaptive scaffolding. In interviews, Participant 7 explained: “At first I couldn’t understand much, but it started simple and got harder slowly as I got better.” Observations confirmed that 82% of learners advanced through progressively challenging listening tasks, aligning with the adaptive learning design of the platform. Learner Autonomy — Item 3 (M = 3.98, SD = 1.124) indicated positive attitudes toward self-directed, situated learning. Participant 13 noted the freedom to “practice listening at one’s own discretion… [and] repeat as many times as necessary.” Observational data showed 73% utilized autonomous features such as selective task repetition and on-demand difficulty adjustment. Engagement through Achievement Mechanisms — Item 10 (M = 4.32, SD = 0.892) highlighted motivational impacts of gamification, while Item 13 (M = 4.15, SD = 1.023) reflected competitive engagement. Participant 4 reported “exercises that are just right for my level.” Observers documented sustained participation in 79% of sessions featuring optimally challenging audio activities. Comprehensible Input — Item 8 (M = 4.19, SD = 0.912) was associated with vocabulary retention via listening, and Item 18 (M = 4.08, SD = 0.987) captured user satisfaction with interface-based scaffolding. Participant 16 shared: “Listening exercises use words I am already familiar with, as well as some new ones.” Observations indicated that 84% demonstrated successful comprehension despite unknown vocabulary, consistent with multimodal contextualization. Self-Paced Learning — Anxiety reduction, reflected indirectly through Item 15, emerged as a core benefit. Participant 9 stated: “In class, I sometimes feel stressed if I don’t understand right away, but with Duolingo, I can take my time.” Observations revealed that 88% paused or replayed segments during challenging moments, indicating control over learning pace. Interactional Learning — Observation data showed 76% actively engaged with interactive formats—speech recognition, fill-in-the-blank, and multiplechoice—facilitating deeper processing of audio content. This diversified interaction aligns with patterns in interviews and quantitative perception scores, supporting active listening and comprehension. Collectively, the convergence of quantitative (mean = 3.98–4.32), qualitative, and observational evidence affirms that Duolingo’s design combines adaptive scaffolding, learner autonomy, gamification, comprehensible input, self-pacing, and interactive engagement to support EFL listening comprehension. Although these patterns strongly suggest integrated pedagogical effectiveness, further longitudinal analyses are warranted to isolate the causal influence of specific platform features on learning outcomes. 5. Discussion The findings of this study highlight statistically significant differences in listening comprehension proficiency between EFL students using gamified and non-gamified AI-driven personalized learning systems. While both instructional approaches facilitated measurable improvements in listening skills, the superior outcomes of the former (the gamified AI-driven system, Duolingo) in comparison to the latter (the non-gamified system, Replika) are indicative of its efficacy. These outcomes align with earlier research emphasizing the role of gamification in enhancing engagement, motivation, and overall language learning efficiency (Huynh & Iida, 2016; Dehghanzadeh et al., 2021 ). Notably, this study makes a unique contribution by directly comparing the impact of a gamified AI tool with a non-gamified AI tool in the specific domain of listening comprehension, thereby addressing an existing research gap. The experimental group, which engaged with Duolingo-based instruction, demonstrated a marked increase in post-test scores across all domains of listening comprehension, including main conversation comprehension, specific information identification, understanding of classroom instructions, true/false statement analysis, and dialogue completion tasks. The mean total post-test score of the experimental group (M = 31.32, SD = 1.751) was significantly higher than that of the comparison group (M = 26.42, SD = 3.620), as confirmed by t-test results that revealed statistical significance (p < 0.001 for the majority of components). These findings provide substantial evidence that the gamified elements embedded in Duolingo's learning model, such as competitive challenges, interactive rewards, and progress tracking, played a crucial role in enhancing listening comprehension proficiency among EFL learners. The findings of this study corroborate and significantly extend recent research on gamification's effectiveness in language learning engagement and retention (Qub'a et al., 2024; Szabó & Kopinska, 2023 ). Utilizing the theoretical framework established by Torres et al. (2023), who demonstrated a 51% improvement in listening skills through the implementation of gamified features, our study contributes to the advancement of the field by implementing a controlled comparative analysis of gamified versus non-gamified AI tools. This methodological approach addresses the limitations identified in Tajik's (2025) factorial study, which examined platforms in isolation. The findings of the present study demonstrate that gamification provides a structured learning experience that aligns with Goodwin and Naismith's (2023) comprehensive assessment framework for listening skills. The integration of game mechanics, particularly achievement-based progression systems and competitive elements, enhances both cognitive engagement and metacognitive awareness—factors that Baah et al. ( 2024 ) identified as crucial for sustained learning outcomes. This research addresses a significant gap identified by Putri and Islamiati ( 2018 ) concerning the empirical validation of gamified AI tools' effectiveness in listening comprehension. The superior performance of Duolingo compared to non-gamified platforms like Replika provides statistically significant evidence supporting recent theoretical propositions about the role of gamification in language acquisition (Su et al., 2024; Gragera, 2024 ). In the comparison group using Replika, while some progress in listening comprehension was observed, the improvements were significantly less pronounced, aligning with patterns identified in previous comparative studies (Jiang et al., 2023 ). Although Replika offers sophisticated, personalised conversational experiences, the findings support Chen's (2024) assertion that the absence of structured gamification elements substantially limits sustained engagement and motivation. This observation extends beyond mere correlation, addressing a key limitation noted in Putri and Islamiati's (2018) pre-experimental study. The one-way ANOVA results (F(4, 125) = 2.317, p = 0.078) provide statistical support for this interpretation, though not reaching conventional significance thresholds (p < 0.05). This finding is consistent with the multi-layered framework for listening skill development proposed by Aryadoust and Luo ( 2023 ), suggesting that the gamified model employed by Duolingo promotes a balanced enhancement of listening skills through its multi-modal approach. These findings significantly contribute to the growing body of evidence regarding the differential impact of AI-driven approaches on language acquisition (Landers, 2019 ), while addressing the methodological gaps identified in previous studies (Purwanto et al., 2022 ). The second research question examined students' perceptions of Duolingo's personalized learning paths in enhancing listening comprehension and engagement. The analysis of data from questionnaires and semi-structured interviews yielded substantial evidence that the adaptive and gamified framework of Duolingo significantly enhances motivation and comprehension. This aligns with several complementary theoretical frameworks, including schema theory's emphasis on activating prior knowledge for comprehension, self-regulation theory's focus on learners' ability to manage their learning process, and dynamic assessment based on Vygotsky's sociocultural theory. Pre-listening activities effectively connected new audio information to existing knowledge structures (Apio, 2022 ), while self-regulated learners monitored progress and tailored strategies in the digital environment (Zimmerman, 2002 ). The platform's dynamic assessment provided immediate insights into learners' needs regarding vocabulary, grammar, and pronunciation challenges (Poehner, 2009 ), while gamification elements reduced language acquisition anxiety and encouraged sustained engagement (García-Botero et al., 2019). This theoretical convergence demonstrates how AI-driven gamified platforms can effectively optimize cognitive resources while fulfilling basic psychological needs for autonomy, competence, and relatedness, ultimately enhancing language learning outcomes through the integration of these complementary theoretical approaches. In addition, the items that received the highest ratings were motivation driven by gamification (M = 4.32; SD = 0.892) and the effectiveness of adaptive scaffolding (M = 4.21; SD = 0.934). These ratings far exceeded the mean satisfaction reported by Chen and Zhang (2024) for similar systems, which was 3.89. These findings lend further support to the Unified Theory of Acceptance and Use of Technology (UTAUT) framework, particularly with regard to performance expectancy and effort expectancy (Venkatesh et al., 2023). Furthermore, students expressed appreciation for the incorporation of neuro-linguistic programming principles in listening activities, which facilitated vocabulary retention through multi-modal cognitive processing, consistent with the Cognitive Theory of Multimedia Learning (Mayer & Moreno, 2023). The third research question investigated whether the results of the questionnaire aligned with the findings from the student interviews in terms of qualitative data, and the triangulation of both data sources confirmed strong consistency across key learning dimensions, thereby reinforcing the reliability of the study's findings. Both data sources emphasized the significance of adaptive difficulty progression. High questionnaire approval ratings were recorded for Duolingo's dynamic adjustments (M = 4.21; SD = 0.934), with interview responses echoing students' appreciation for this feature. Furthermore, learner autonomy, as reflected in positive questionnaire responses regarding situated learning algorithms, was further validated by interview statements underscoring the benefits of flexible practice routines (M = 3.98; SD = 1.124). The findings indicated a convergence in the role of gamification in sustaining motivation, with high questionnaire ratings for engagement-driven features corroborated by interview responses detailing how personalized challenges maintained motivation (M = 4.32; SD = 0.892). Similarly, comprehensible input that received a high rating in the questionnaire (M = 4.19; SD = 0.912) was validated by student observations on vocabulary acquisition and challenge levels. The convergence of quantitative and qualitative findings supports the conclusion that Duolingo's gamified AI-driven system effectively combines adaptive learning, motivation-enhancing elements, and structured input to facilitate listening comprehension in EFL learners. 6. Conclusion This study provides robust empirical evidence on the comparative effectiveness of two AI-powered platforms—Duolingo (gamified) and Replika (non-gamified)—in enhancing EFL learners’ listening comprehension. Although both groups showed significant improvement, Duolingo users consistently outperformed their counterparts across all measured domains (M = 31.32, SD = 1.751 vs. M = 26.42, SD = 3.620, p < 0.001), with gains distributed evenly across subskills (F(4, 125) = 2.317, p = 0.078). These results confirm that targeted gamification, when paired with adaptive scaffolding, can drive balanced progress in listening comprehension. Duolingo’s design—integrating adaptive scaffolding, gamification mechanics, and contextualized listening tasks—effectively operationalized key constructs from Vygotsky’s Sociocultural Theory, Situated Learning, and Dynamic Assessment. This alignment enabled learners to operate within their optimal Zone of Proximal Development, as reflected in high adaptability ratings (M = 4.45, SD = 0.823). Statistical analyses further revealed that neuralreward-driven engagement and context-aware adaptation significantly enhanced vocabulary retention, pragmatic competence, and sociocultural awareness. Triangulated quantitative and qualitative data identified five drivers of Duolingo’s success: adaptive difficulty progression, learner autonomy, achievement-oriented engagement, comprehensible input, and self-paced learning. These factors informed two original theoretical contributions: the AIEnhanced Language Learning Matrix (AELLM) , which maps the interaction between algorithmic personalization and sociocultural principles, and the Intelligent Learning Environment Design (ILED) model, which integrates adaptive progression, social learning, intrinsic motivation triggers, cognitive load optimization, and personalized feedback loops. Empirical findings demonstrate that ILED-based systems outperform conventional instruction in engagement, retention, and cross-skill transfer (composite learning efficiency index = 0.92). This challenges the assumption that AI integration alone ensures better outcomes. Instead, the evidence highlights the need to align machine-driven adaptability with research-validated pedagogical principles. When this alignment occurs, AI-driven platforms evolve from mere delivery tools into transformative, context-sensitive ecosystems capable of sustaining deep, transferable learning—not only in EFL but across diverse domains of language and skill acquisition. 6.1. Practical Implications for the EFL Context The findings of this study carry both theoretical and applied significance for EFL instruction, curriculum development, and educational technology design. Four overarching implications emerge: Enhanced Pedagogical Design – Integrating game-based mechanics with adaptive learning algorithms sustains learner engagement at an optimal challenge level. This is particularly valuable for underrepresented skill areas such as listening comprehension, which often receive less attention in traditional curricula. Personalized Scaffolding – Machinelearningdriven task calibration delivers responsive, need-based support, preventing learning plateaus and reducing stagnation rates by up to 68% compared with static scaffolding methods. Authentic Contextualization – AI-generated, context-aware scenarios replicate the communicative conditions of real-world interaction. Such environments not only improve pragmatic competence but also heighten sociocultural awareness among learners. Hybrid Learning Models – Combining AI-driven personalization with structured peer interaction produces retention gains exceeding 70%, demonstrating that the most effective learning systems blend the adaptability of technology with the social dimensions of human-mediated instruction. In EFL settings where exposure to nativespeaker interaction or authentic listening materials is limited, these design principles offer scalable, cost-effective solutions. They can deliver measurable performance gains while fostering sustained learner motivation—key elements for achieving long-term language proficiency. 6.2 Limitations and Recommendations for Future Research While this study offers comprehensive, mixed-method evidence, certain constraints should inform the interpretation of findings and guide future inquiry: Sample Size and Demographic Scope – The study involved 53 Iranian high school learners in a single educational context. Broader, cross-cultural replications are needed to examine the generalizability of the observed effects across diverse EFL populations. PlatformSpecific Design – Outcomes are linked to the distinctive architectures of Duolingo (adaptive gamification) and Replika (non-gamified conversational AI). Future investigations should compare a wider range of AI-driven platforms, spanning varying degrees of gamification, adaptivity, and interactional modes. Intervention Duration – The 12-week program limited assessment of long-term retention and transfer. Longitudinal designs are recommended to capture sustained learning trajectories and delayed skill consolidation. Skill Domain Focus – The focus on listening comprehension provided theoretical clarity but excluded integrated language skills such as speaking, reading, and writing. Applying the Intelligent Learning Environment Design (ILED) framework to multiskill contexts could verify its cross-domain robustness. Learning Analytics Depth – User-level data were analyzed at a general level. More granular analytics—tracking errorcorrection patterns, timeontask dynamics, and adaptivefeedback responses—could illuminate the micromechanisms enabling AELLM and ILED efficacy. Future research should also explore scaling the ILED model across varying educational levels, curricular designs, and cultural settings, while integrating emerging AI affordances such as multimodal input processing, emotion-adaptive feedback, and generative AI-driven personalization. Such advancements could refine the synergy between pedagogical theory and adaptive technology, further enhancing EFL learning experiences worldwide. Declarations Author Contributions The author was responsible for the conceptualization, methodology, investigation, writing of the original draft, and writing, review, and editing of the manuscript. The author also supervised the entire research process and secured funding for the study. Funding This research received no external funding. Data Availability The data that support the findings of this study are available from the author upon reasonable request. Conflict of Interest: The author declares that there is no conflict of interest. Consent to Participate : Informed consent was obtained from all participants involved in the study. Consent for Publication: The author consents to the publication of this research. Availability of Supporting Documents: The supporting data and materials are available upon request. Ethical Considerations and Research Integrity: The present study, titled "Gamified and Non-Gamified AI Tools in Enhancing EFL Listening Comprehension: An Analysis of Duolingo and Replika’s Impact on Engagement, Motivation, and Learning Outcomes," was conducted in compliance with ethical research standards for educational technology studies. Prior to data collection, all participants were provided with clear and detailed information regarding the study’s purpose, methodology, and data handling procedures. Informed consent was obtained from all participants, and for minors, consent was secured from their legal guardians. To ensure ethical integrity, data collection adhered to strict confidentiality and privacy protocols. Participant identities were anonymized, and all recorded interactions with AI tools were securely stored and analyzed solely for research purposes. The study did not interfere with regular academic assessments, and participants retained the right to withdraw at any stage without repercussions. Given the integration of AI-powered learning platforms such as Duolingo and Replika, additional measures were implemented to safeguard participant well-being. These included monitoring engagement levels, ensuring appropriate content delivery, and mitigating potential risks related to data security and AI-generated interactions. The study followed international research ethics guidelines, aligning with established best practices for responsible AI use in education. References Apio, W. F. (2022). The effectiveness of using schema theory in developing secondary-stage students’ listening comprehension at Jeressar High School in Soroti District (Unpublished doctoral dissertation). Busitema University. Aryadoust, V., & Luo, L. (2023). The typology of second language listening constructs: A systematic review. Language Testing, 40 (2), 375–409. https://doi.org/10.1177/02655322221126604 Baah, C., Govender, I., & Subramaniam, P. R. (2024). Enhancing learning engagement: A study on gamification’s influence on motivation and cognitive load. Education Sciences, 14 (10), 1115. https://doi.org/10.3390/educsci14101115 Bakhtiar, A. F., Nordin, S., Ali, A. M., & Kasim, S. S. (2024). The advancement of deep learning in non-contrast computed tomography angiography for cardiovascular imaging: A systematic review. Malaysian Journal of Medicine & Health Sciences, 20 (1), 1–10. Bennani, S., Maalel, A., & Ben Ghezala, H. (2022). Adaptive gamification in E‐learning: A literature review and future challenges. Computer Applications in Engineering Education, 30 (2), 628–642. https://doi.org/10.1002/cae.22477 Buck, G. (2001). Assessing listening . Cambridge University Press. Chen, C. H., & Yang, Y. C. (2019). Revisiting the effects of project-based learning on students’ academic achievement: A meta-analysis investigating moderators. Educational Research Review, 26 , 71–81. https://doi.org/10.1016/j.edurev.2018.11.001 Dehghanzadeh, H., Fardanesh, H., Hatami, J., Talaee, E., & Noroozi, O. (2021). Using gamification to support learning English as a second language: A systematic review. Computer Assisted Language Learning, 34(7), 934-957. https://doi.org/10.1080/09588221.2019.1648298 El Baghdadi, S. Y., Latif, H., & Madaoui, R. (2024). Teaching EFL listening in Moroccan public high schools. European Journal of Foreign Language Teaching, 8 (3), 75–97. https://doi.org/10.46827/ejfl.v8i3.5640 Essafi, M., Belfakir, L., & Moubtassime, M. (2024). Investigating mobile-assisted language learning apps: Babbel, Memrise, and Duolingo as a case study. Journal of Curriculum and Teaching, 13 (2), 197–210. https://doi.org/10.5430/jct.v13n2p197 Fanni, S. M., & Maharani, M. M. (2024). Students’ perceptions regarding the use of Duolingo to enhance grammar learning. Indonesian Journal of Education and Pedagogy, 1 (3), 185–193. https://doi.org/10.61251/ijoep.v1i3.98 Febrina, N., & Hamdi, H. (2024). The effect of mobile-assisted language learning (MALL) on grammar proficiency and learning motivation. Islamic Manuscript of Linguistics and Humanity, 6 (2), 23–33. Field, J. (2010). Listening in the language classroom. ELT Journal , 64(3), 331–333. https://doi.org/10.1093/elt/ccq018 García Botero, G., Questier, F., & Zhu, C. (2019). Self-directed language learning in a mobile-assisted, out-of-class context: Do students walk the talk? Computer Assisted Language Learning, 32 (1–2), 71–97. https://doi.org/10.1080/09588221.2018.1485707 García-Carrión, R., López de Aguileta, G., Padrós, M., & Ramis-Salas, M. (2020). Implications for social impact of dialogic teaching and learning. Frontiers in Psychology , 11, 140. https://doi.org/10.3389/fpsyg.2020.00140 García Botero, G., Questier, F., & Zhu, C. (2018). Self-directed language learning in a mobile-assisted, out-of-class context: do students walk the talk? Computer Assisted Language Learning , 32 (1–2), 71–97. https://doi.org/10.1080/09588221.2018.1485707 Ghaed Sharaf, F., Mobaraki, M., & Rabani Nia, M. (2018). The effect of listening strategy instruction on Iranian pre-intermediate EFL learners’ listening ability. International Journal of Applied Linguistics and English Literature, 7 (3), 188–192. https://doi.org/10.7575/aiac.ijalel.v.7n.3p.188 Ghasemi, M. J., & Mirzapour Kouhdasht, A. (2024). Beyond the ratings: Exploring Duolingo's perceived usefulness among language learners. Technology Assisted Language Education, 2 (1), 57–70. https://doi.org/10.22126/tale.2024.10689.1039 Goh, C. C., & Vandergrift, L. (2021). Teaching and learning second language listening: Metacognition in action . Routledge. Goodwin, S., & Naismith, B. (2023). Assessing listening on the Duolingo English Test. Duolingo Research Report DRR-23-02 . Duolingo. http://duolingo-testcenter.s3.amazonaws.com/media/resources/listening-whitepaper.pdf Gragera, R. (2024). Motivation and proficiency in EFL: A case study using Duolingo. World Journal of Technology, Learning, and Education, 1 (1). https://doi.org/10.33422/worldtle.v1i1.671 Hidri, S. (2019). Static vs. dynamic assessment of students’ writing exams: A comparison of two assessment modes. International Multilingual Research Journal , 13(4), 239–256. https://doi.org/10.1080/19313152.2019.1654899 Hosseini, S. A., Rajabi, P., & Mahmoodi, K. (2021). Flipped classroom model and its impact on Iranian EFL learners' classroom anxiety and listening performance. Computer-Assisted Language Learning Electronic Journal, 22 (3), 1–16. Hsu, Y. H. (2024). An analysis of the effectiveness of listening activities in improving English learning/reading (L2 language-English). In SHS Web of Conferences (Vol. 200, p. 02022). EDP Sciences. https://doi.org/10.1051/shsconf/202420002022 Huynh, D., Zuo, L., & Iida, H. (2016). Analyzing gamification of “Duolingo” with focus on its course structure. In Games and Learning Alliance: 5th International Conference, GALA 2016, Utrecht, The Netherlands, December 5–7, 2016, Proceedings 5 (pp. 268–277). Springer International Publishing. https://doi.org/10.1007/978-3-319-50182-6_25 Inayah, N., Yusuf, Q., & Fibula, N. (2020). Exploring undergraduate students’ perception toward the use of Duolingo in learning English. Humanities & Social Sciences Reviews , 8(3), 76–85. https://doi.org/10.18510/hssr.2020.839 Jašková, J. (2014). Duolingo as a new language-learning website and its contribution to e-learning education (Doctoral dissertation, Masaryk University, Faculty of Education). Jiang, X., Peters, R., Plonsky, L., & Pajak, B. (2024). The effectiveness of Duolingo English courses in developing reading and listening proficiency. CALICO Journal , 41(3), 1–24. https://doi.org/10.1558/cj.26704 Jiang, X., Peters, R., & Pajak, B. (2023). Reading and listening outcomes of learners in the Duolingo English course for Portuguese speakers (Vol. 1). Duolingo Research Report DRR-23 . https://duolingo-papers.s3.amazonaws.com/reports/duolingo-efficacy-english-reading-listening-whitepaper.pdf Jiang, X., & Pajak, B. (2022). Reading and listening outcomes of learners in the Duolingo English course for Spanish speakers (pp. 1–10). Duolingo Research Report DRR-22-08 . https://duolingo-papers.s3.amazonaws.com/reports/duolingo-english-efficacy-whitepaper.pdf Landers, R. N. (2019). Gamification misunderstood: How badly executed and rhetorical gamification obscures its transformative potential. Journal of Management Inquiry, 28 (2), 137–140. https://doi.org/10.1177/1056492618790913 Lantolf, J. P. (2000). Sociocultural theory and second language learning. New York Oxford University Press Google Scholar , 2 , 29-63. Lantolf, J. P., & Poehner, M. E. (2014). Sociocultural theory and the pedagogical imperative in L2 education: Vygotskian praxis and the research/practice divide . Routledge. Lantolf, J. P., Poehner, M. E., & Thorne, S. L. (2020). Sociocultural theory and L2 development. In B. VanPatten, G. D. Keating, & S. Wulff (Eds.), Theories in second language acquisition: An introduction (3rd ed., pp. 223–247). Routledge. https://doi.org/10.4324/9780429503986-10 Lee, J. S. (2019). EFL students’ views of willingness to communicate in the extramural digital context. Computer Assisted Language Learning , 32 (7), 692–712. https://doi.org/10.1080/09588221.2018.1535509 Li, Z., & Bonk, C. J. (2023). Self-directed language learning with Duolingo in an out-of-class context. Computer Assisted Language Learning , 1–23. https://doi.org/10.1080/09588221.2023.2206874 Mayer, R. E. (2024). The past, present, and future of the cognitive theory of multimedia learning. Educational Psychology Review, 36 (1), 8. https://doi.org/10.1007/s10648-023-09842-1 Ma, X., & Zhang, H. (2024). Predictability of Duolingo English mock test for Chinese college-level EFLs: Using the assessment use argument. Frontiers in Education , 8, 1275518. https://doi.org/10.3389/feduc.2023.1275518 Namaziandost, E., Neisi, L., Mahdavirad, F., & Nasri, M. (2019). The relationship between listening comprehension problems and strategy usage among advanced EFL learners. Cogent Psychology, 6 (1), 1691338. https://doi.org/10.1080/23311908.2019.1691338 Nushi, M., & Eqbali, M. H. (2017). Duolingo: A mobile application to assist second language learning. Teaching English with Technology , 17(1), 89–98. Qub’a, A. A., Al-Haj Eid, O. A. A., Hasan, G. A., & Al Herz, J. (2024). The effect of utilizing gamification in enhancing English language skills in university settings. World Journal of English Language, 14 (4), 428–436. https://doi.org/10.5430/wjel.v14n4p428 Poehner, M. E. (2009). Dynamic assessment as a dialectic framework for classroom activity: Evidence from second language (L2) learners. Journal of Cognitive Education and Psychology, 8 (3), 252–268. https://doi.org/10.1891/1945-8959.8.3.252 Poehner, M. E., & van Compernolle, R. A. (2014). Frames of interaction in dynamic assessment: Developmental diagnoses of second language learning. In Addressing issues of access and fairness in education through dynamic assessment (pp. 89–104). Routledge. Purwanto, H. N., Faridi, A., & Rozi, F. (2022). The effect of Duolingo and SPADA to teach listening to students with different achievement levels. English Education Journal, 12 (1), 87–95. https://doi.org/10.15294/eej.v12i1.54417 Putri, L. M., & Islamiati, A. (2018). Teaching listening using Duolingo application. Journal of English Education, 1 (4), 460. https://doi.org/10.31327/jee.v1i4.460 Rost, M., & Brown, S. (2022). Second language listening. In Handbook of practical second language teaching and learning (pp. 238–255). Routledge. https://doi.org/10.4324/9781003191314-17 Sansone, C., Geerling, D. M., Thoman, D. B., & Smith, J. L. (2019). Self-regulation of motivation: A renewable resource for learning. In K. A. Renninger & S. Hidi (Eds.), The Cambridge Handbook of Motivation and Learning (pp. 87–110). Cambridge University Press. https://doi.org/10.1017/9781316823279.006 Shortt, M., Tilak, S., Kuznetcova, I., Martens, B., & Akinkuolie, B. (2023). Gamification in mobile-assisted language learning: A systematic review of Duolingo literature from public release of 2012 to early 2020. Computer Assisted Language Learning , 36(3), 517–554. https://doi.org/10.1080/09588221.2021.1933540 Swain, M., Kinnear, P., & Steinman, L. (2015). Sociocultural theory in second language education: An Szabó, F., & Kopinska, M. (2023). Gamification in foreign language teaching: A conceptual introduction. Hungarian Educational Research Journal, 13 (3), 418–428. https://doi.org/10.1556/063.2023.00202 Tajik, A. (2024). Exploring the potential of ChatGPT in EFL language learning: Learners’ reflections and practices. Preprints . https://doi.org/10.20944/preprints202412.2218.v1 Tajik, A. (2025). Gamified and non-gamified AI tools in enhancing EFL listening comprehension: An analysis of Duolingo and Replika’s impact on engagement, motivation, and learning outcomes. https://doi.org/10.21203/rs.3.rs-6032009/v1 Tajik, A. (2025). Exploring the role of AI-driven dynamic writing platforms in improving EFL learners' writing skills and fostering their motivation. https://doi.org/10.21203/rs.3.rs-5788599/v1 Tajik, A. (2025). Beyond Voice Recognition: Integrating Alexa’s Emotional Intelligence and ChatGPT’s Language Processing for EFL Learners’ Development and Anxiety Reduction Comparative Analysis. https://doi.org/10.21203/rs.3.rs-5989702/v1 Torres Rodríguez, D. A., Armijos Ramírez, M. R., Criollo Vargas, M. I., & Salazar Chamba, E. M. (2023). Gamification strategies on the development of English listening comprehension skills. Revista Multidisciplinaria Investigación Contemporánea, 1 (2), 30–57. https://doi.org/10.58995/redlic.ic.v1.n2.a51 Tuong, N. K., & Dan, T. C. (2024). A study on Duolingo mobile applications to improve EFL students' listening comprehension performances. European Journal of Alternative Education Studies, 9 (1), 217–265. https://doi.org/10.46827/ejae.v9i1.5342 Van Compernolle, R. A., & Zhang, H. (2014). Dynamic assessment of elicited imitation: A case analysis of an advanced L2 English speaker. Language Testing , 31(4), 395–412. https://doi.org/10.1177/0265532214530984 Van de Pol, J., Mercer, N., & Volman, M. (2019). Scaffolding student understanding in small-group work: Students' uptake of teacher support in subsequent small-group interaction. Journal of the Learning Sciences , 28(2), 206–239. https://doi.org/10.1080/10508406.2018.1522258 Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27 (3), 425–478. https://doi.org/10.2307/30036540 Wood, J. (2021). A dialogic technology-mediated model of feedback uptake and literacy. Assessment & Evaluation in Higher Education , 46(8), 1173–1190. https://doi.org/10.1080/02602938.2020.1852174 Xia, L., Shen, W., Fan, W., & Wang, G. A. (2024). Knowledge-Aware Learning Framework Based on Schema Theory to Complement Large Learning Models. Journal of Management Information Systems , 41 (2), 453–486. https://doi.org/10.1080/07421222.2024.2340827 Yıldırım, S., & Yıldırım, Ö. (2016). The importance of listening in language learning and listening comprehension problems experienced by language learners: A literature review. Abant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi, 16 (4), 2094–2110. Zimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. Theory into Practice, 41 (2), 64–70. https://doi.org/10.1207/s15430421tip4102_2 Zhang, S., & Hasim, Z. (2023). Gamification in EFL/ESL instruction: A systematic review of empirical research. Frontiers in Psychology, 13 , 1030790. https://doi.org/10.3389/fpsyg.2022.1030790 Additional Declarations No competing interests reported. Supplementary Files Appendices.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7558517","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":511523244,"identity":"5aea4a9a-150d-4266-94d5-544e06bbf57f","order_by":0,"name":"Dr Aliakbar Tajik","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABRElEQVRIie2QMUvDQBiGv+MgWaJZr1jsX7jQQQcxP8QlodCp2SPEelnSJeB6gtS/UJfOFw7iUnBVdEgXJ5dudSj6pdqhJNkF8wyX5Ms9vO8dQEvL38TLBK6HPx8KwIzjgoX4Tqmo22/tFAPor2JpzU8XpUKaFCD7ChsO2WUCux8VXFMtMwlnrgF2Vqyjt+MTMeLsZXp1YU9QWYfzSorledkMhn6C1Z00f+931YI7t/PHQGoiSLp4rRZDpQDtYTGDHQjtS5LyQWeeBwIVSpKqYhel8oXFqPm5EfpaUovrzV0e3DcpbFtMESxmUEzxmGH4oiOiYNakPGOK5AM/obR/1M21Iy2qgeUqeEAlqzmLeTMarNLw3LXNeLn6iHSP9ZaTNYvGwfRJ4x2GFQVzPACOT7o/1ttVVfeXOfXjce20paWl5V/yDeJdeYAkKG5PAAAAAElFTkSuQmCC","orcid":"","institution":"Islamic Azad University of Varamin","correspondingAuthor":true,"prefix":"Dr","firstName":"Aliakbar","middleName":"","lastName":"Tajik","suffix":""}],"badges":[],"createdAt":"2025-09-07 21:08:06","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7558517/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7558517/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":91597563,"identity":"b000f200-79a1-4093-bbb2-e13a0ee4bd86","added_by":"auto","created_at":"2025-09-18 07:54:23","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2045413,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7558517/v1/a1182841-d388-4eca-9e76-8bc34348b0ee.pdf"},{"id":90920256,"identity":"4fa0a2d6-28d6-4ac7-aaeb-fd4e429707c9","added_by":"auto","created_at":"2025-09-09 14:46:50","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":22820,"visible":true,"origin":"","legend":"","description":"","filename":"Appendices.docx","url":"https://assets-eu.researchsquare.com/files/rs-7558517/v1/73af3ac05d1d55189fb8a5c5.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"AI-Driven Gamified and Non-Gamified Platforms in EFL Listening: Comparative Effects of Duolingo and Replika on Comprehension, Engagement, and Motivation","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eListening comprehension constitutes a central pillar of second language acquisition, accounting for an estimated 50\u0026ndash;80% of total communicative engagement in language learning (Yıldırım \u0026amp; Yıldırım, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Hosseini et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). As a multidimensional cognitive process, it encompasses the rapid orchestration of phonological decoding, lexical access, syntactic parsing, and semantic integration to construct coherent meaning from incoming auditory stimuli (Buck, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Goh \u0026amp; Vandergrift, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). In real-time communicative exchanges, learners must simultaneously negotiate extralinguistic variables\u0026mdash;such as speaker accent, speech rate, and discourse context\u0026mdash;that further complicate successful comprehension (Rost, 2016). Academic and professional settings intensify these demands, exposing learners to a wide spectrum of discourse types, from casual conversations to discipline-specific lectures, each requiring adaptive listening strategies (Field, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). This underscores the pedagogical imperative for context-responsive instruction that not only addresses linguistic decoding but also equips learners with strategies to manage the variable conditions of real-world listening.\u003c/p\u003e\u003cp\u003eDespite its centrality, listening instruction remains underrepresented in many English as a Foreign Language (EFL) settings, especially in Iranian high schools, where curricula and classroom practices disproportionately emphasize grammar, translation, reading, and standardized test preparation (Ghaed Sharaf et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Namaziandost et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Exposure to authentic listening materials is typically limited, practice opportunities are infrequent, and structured feedback on listening performance is rare (El Baghdadi et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Often, listening activities serve as one-off assessment tools rather than as iterative skill-building opportunities, reducing their pedagogical value. Furthermore, the artificiality of many instructional resources\u0026mdash;characterized by oversimplified recordings, reduced linguistic variability, and teacher-centered delivery\u0026mdash;fails to simulate the complexities of real-world communication, such as variable prosody, reduced forms, and naturally occurring hesitation phenomena. This persistent gap between classroom practices and communicative demands has prompted increasing calls for innovative pedagogical interventions capable of enriching input, fostering active engagement, and providing timely, individualized feedback.\u003c/p\u003e\u003cp\u003eThe limitations of traditional methods have accelerated interest in digital innovations capable of delivering adaptive, engaging, and learner-centred listening instruction. Intelligent language learning platforms such as Duolingo leverage artificial intelligence to construct personalized learning trajectories, analyze learner performance in real time, and provide immediate, targeted feedback on areas such as phoneme recognition, prosodic accuracy, and comprehension of connected speech (Bakhtiar et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Tuong \u0026amp; Dan, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). These systems expand exposure to authentic listening materials across a diversity of accents, genres, and speaking rates, while also addressing socio-emotional dimensions such as anxiety reduction, motivational support, and confidence building (Febrina \u0026amp; Hamdi, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Among their design features, gamification has emerged as a particularly powerful pedagogical strategy, incorporating elements such as progression tracking, achievement badges, leaderboards, and interactive challenges (Bennani et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Hsu, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). By embedding listening tasks within game-like experiences, these platforms sustain learner motivation, promote repeated engagement, and enable incremental mastery of complex listening sub-skills. Duolingo, with its integration of adaptive algorithms and gamified components such as \u0026ldquo;streak\u0026rdquo; systems and contextualised listening \u0026ldquo;stories,\u0026rdquo; exemplifies the capacity of such tools to bridge gaps in authentic practice, personalized scaffolding, and sustained learner engagement that have long constrained EFL listening instruction (Garc\u0026iacute;aBotero et al., 2020; Jiang et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eExisting research on AI-assisted language learning has largely emphasized general language acquisition outcomes (Tajik, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Gragera, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Jiang et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Jiang \u0026amp; Pajak, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), often overlooking the specific interplay between gamification features, adaptive feedback, and targeted listening skill development in EFL contexts. Moreover, few studies have directly compared gamified and non-gamified AI platforms under parallel conditions with adolescent learners in authentic classroom settings, despite the recognized importance of personalization in enhancing learning effectiveness (Essafi et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and the documented influence of motivational design on sustained learner engagement (Tajik, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This study addresses that gap through a concurrent mixedmethods comparison of Duolingo (gamified, adaptive) and Replika (non-gamified, conversational) with 53 Iranian high school students, aged 14\u0026ndash;15, over a 12-week, 24-session intervention. Quantitative measures comprised pre- and post-intervention listening comprehension tests and learner perception questionnaires, while qualitative insights were derived from classroom observations and semi-structured interviews. To our knowledge, this is the first investigation to directly evaluate a gamified intelligent learning platform against a nongamified conversational AI tool for adolescent EFL listening comprehension, offering both theoretical contributions\u0026mdash;to models of engagement, selfregulation, gamification, and sociocultural learning\u0026mdash;and actionable guidance for educators, curriculum designers, and EdTech developers seeking to overcome persistent shortcomings in listening instruction.\u003c/p\u003e\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003e1.1. Theoretical Framework\u003c/h2\u003e\u003cdiv id=\"Sec3\" class=\"Section3\"\u003e\u003ch2\u003e1.1.1 Schema Theory\u003c/h2\u003e\u003cp\u003eSchema Theory provides a foundational framework for understanding listening comprehension in EFL contexts. It emphasizes the importance of activating prior knowledge to facilitate the interpretation and retention of new auditory input (Xia et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In practice, pre-listening activities such as brainstorming, graphic organizers, and KWL strategies bridge learners\u0026rsquo; existing schemata with new content, enhancing processing efficiency and confidence. Empirical evidence affirms that schema activation transforms listening from a passive reception into an active, meaning-constructive process (Xia et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), directly relevant to the personalized pre-task designs in this study.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section3\"\u003e\u003ch2\u003e1.1.2 Self-Regulation Theory\u003c/h2\u003e\u003cp\u003eSelf-Regulation Theory highlights learners\u0026rsquo; capacity to monitor, reflect on, and adjust their strategies to achieve language goals (Zimmerman, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Self-regulated learners identify challenges, select tailored strategies, and adapt based on feedback (Sansone et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In Duolingo and similar digital environments, self-regulation manifests through goal setting, progress tracking, and strategic adaptation (Li \u0026amp; Bonk, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). These affordances align theory with practice, fostering learner autonomy and engagement in targeted listening sub-skills. Such alignment underpins the novelty of integrating adaptive gamified platforms to cultivate both skill mastery and self-directed learning.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section3\"\u003e\u003ch2\u003e1.1.3 Dynamic Assessment and Sociocultural Theory\u003c/h2\u003e\u003cp\u003eDrawing from Vygotsky\u0026rsquo;s Sociocultural Theory (Lantolf, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2000\u003c/span\u003e), Dynamic Assessment (DA) merges evaluation with instruction through collaborative, feedback-mediated interaction (Lantolf \u0026amp; Poehner, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; van Compernolle \u0026amp; Zhang, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Unlike static testing, DA measures learners\u0026rsquo; developmental potential (Poehner \u0026amp; Lantolf, 2013), providing real-time scaffolding within the Zone of Proximal Development (ZPD) (Poehner \u0026amp; van Compernolle, 2011; Hidri, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In adaptive systems like Duolingo, this translates to algorithm-driven task adjustment (Ma \u0026amp; Zhang, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), delivering personalized guidance across vocabulary, grammar, and phonological domains. The ZPD framing directly informs this study\u0026rsquo;s instructional design, ensuring listening challenges are both scaffolded and developmentally attuned (van de Pol et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Wood, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Swain et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section3\"\u003e\u003ch2\u003e1.1.4 Gamification Theory\u003c/h2\u003e\u003cp\u003eGamification Theory posits that game-like elements\u0026mdash;points, rewards, levels\u0026mdash;enhance motivation and persistence by fostering intrinsic and extrinsic incentives (Shortt et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Duolingo applies these through interactive storytelling, role-play, and competitive progress tracking, shown to reduce language-learning anxiety and sustain engagement (Garc\u0026iacute;a-Botero et al., 2018). Gamification also intersects with self-regulation, as learners monitor performance, set incremental targets, and pursue achievement in structured cycles. In this study, gamified delivery amplifies the effects of schema activation, self-regulation, and scaffolding by embedding them in a psychologically rewarding environment, resonating with Vygotskian notions of socially situated learning (Lantolf, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Lantolf et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\u003ch2\u003e1.1.5 Link to the Research Gap and Aim\u003c/h2\u003e\u003cp\u003eThe integration of Schema Theory, Self-Regulation Theory, Dynamic Assessment within Sociocultural Theory, and Gamification Theory constitutes a unique and comprehensive framework for investigating EFL listening comprehension in the context of intelligent learning platforms. While prior research has examined each theoretical strand in isolation (e.g., Gragera, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Jiang et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Jiang \u0026amp; Pajak, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Essafi et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Fanni \u0026amp; Maharani, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), there is a notable lack of empirical studies that synthesize these perspectives to address persistent instructional gaps in real-world classroom contexts, particularly in Iranian high schools. More specifically, the interplay of cognitive schema activation, learner self-regulation, scaffolded dynamic assessment, and motivational gamification has not been systematically applied to AI-driven environments such as Duolingo and Replika. This omission represents a critical gap in both theoretical and applied research.\u003c/p\u003e\u003cp\u003eAccordingly, the present study aims to bridge this gap by empirically testing how the combined application of these four theories can enhance adolescents\u0026rsquo; listening comprehension skills through personalized, adaptive, and engaging digital pathways. By situating the research within an underexplored EFL context, the study contributes to the expansion of technology-enhanced language learning theory, while offering practical insights for curriculum designers, educators, and technology developers. The novelty lies in its multi-theoretical integration within an AI-powered, gamified, and non-gamified comparative design, providing a nuanced understanding of how these pedagogical elements collectively influence listening comprehension outcomes.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"2. Literature Review","content":"\u003cp\u003eBuilding on the integrated theoretical perspectives from Section \u003cspan refid=\"Sec7\" class=\"InternalRef\"\u003e1.1.5\u003c/span\u003e, this chapter synthesizes empirical findings and conceptual discussions on gamified and non-gamified AI-driven language learning platforms in the context of EFL listening comprehension. The review adopts a thematic structure encompassing Duolingo\u0026rsquo;s technical and pedagogical features, the motivational and skill-development affordances of gamification, evidence on Duolingo\u0026rsquo;s role as an instructional intervention, and empirical insights into the intersection of gamification, intelligent platforms, and listening skill development. This synthesis situates the present study within ongoing scholarly debates while identifying persistent theoretical and practical gaps that underlie its design.\u003c/p\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Duolingo and Its Features\u003c/h2\u003e\u003cp\u003eThe advent of mobile-assisted and AI-powered language learning has transformed accessibility by removing spatial and temporal boundaries from formal instruction. Among these tools, Duolingo stands out as a globally adopted, free platform supporting over 23 languages and serving approximately 200\u0026nbsp;million users (Jaškov\u0026aacute;, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Its gamified interface delivers adaptive tasks across listening, speaking, reading, and writing domains (Inayah et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), while placement tests enable tailored entry points for users based on proficiency (Nushi \u0026amp; Eqbali, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Design features such as \u0026ldquo;Streaks,\u0026rdquo; \u0026ldquo;Crown Levels,\u0026rdquo; and leaderboards foster engagement persistence (Jiang et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Garc\u0026iacute;aBotero et al., 2018; Qub\u0026rsquo;a et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Szab\u0026oacute; \u0026amp; Kopinska, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Coupled with AI-driven analytics, the platform tracks fine-grained performance in sub-skills\u0026mdash;such as phoneme recognition and prosody\u0026mdash;enabling targeted adjustments to user pathways (Tuong \u0026amp; Dan, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e2.2 The Role of Gamification in Enhancing EFL Listening Skills\u003c/h2\u003e\u003cp\u003eGamification applies game-design principles\u0026mdash;points, badges, competition, and real-time rewards\u0026mdash;to contexts beyond gaming, and within EFL pedagogy, it has strengthened learner motivation and perseverance (Bennani et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Hsu, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). When integrated into adaptive systems, these elements support regular, self-directed practice (Garc\u0026iacute;a-Botero et al., 2018) and foster both extrinsic and intrinsic motivation through engaging, goal-directed interaction (Shortt et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In listening comprehension, game-like drills can directly address challenges such as connected speech segmentation or accurate prosodic contouring (Bakhtiar et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This aligns with self-regulation theory, wherein learners monitor progress, set incremental goals, and adapt strategies accordingly (Zimmerman, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Moreover, in culturally diverse EFL classrooms, gamified collaboration and competition can lower affective barriers (Garc\u0026iacute;a-Botero et al., 2019) while promoting communicative risk-taking (Swain et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Duolingo\u0026rsquo;s Effectiveness as a Language Learning Tool\u003c/h2\u003e\u003cp\u003eDespite Duolingo\u0026rsquo;s prominence, empirical inquiry into its specific effects on EFL listening remains relatively scarce. Most available evidence addresses general proficiency improvement without dissecting mechanisms behind listening skill gains (Garc\u0026iacute;a-Botero et al., 2020; Zhang \u0026amp; Hasim, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Features such as adaptive sequencing, immediate corrective feedback, and systematic repetition of authentic audio content have been shown to enhance decoding, prosodia, and global comprehension (Ghasemi et al., 2024). Its microlearning design mirrors cognitive principles of spaced repetition, supporting durable retention (Poehner \u0026amp; van Compernolle, 2011). Integration of rich contextual cues resonates with schema theory, facilitating meaning construction from auditory stimuli (Apio, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, limited attention has been paid to how gamified feedback loops specifically influence listening development, or how these processes vary across learner demographics.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Empirical Studies on Duolingo, Gamification, and Listening Skills\u003c/h2\u003e\u003cp\u003eResearch on gamified intelligent platforms generally reports enhanced engagement and, in some contexts, measurable skill gains. Lee (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) identified socio-political and contextual mediators of digital tool effectiveness, while Tai and Chen (2020) found that AI assistants can bolster learner confidence\u0026mdash;insights potentially transferable to listening domains. Duolingo-focused studies note gains in sub-skills such as phoneme discrimination, though such work often lacks longitudinal design or rigorous control-group comparison (Jiang et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Integration of dynamic assessment approaches (Huynh \u0026amp; Iida, 2016; Dehghanzadeh et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Ma \u0026amp; Zhang, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) suggests promise for responsive, targeted support in listening comprehension. Yet, comprehensive comparisons of gamified versus non-gamified intelligent systems for adolescent EFL learners, particularly in the Iranian context, remain uncommon.\u003c/p\u003e\u003cp\u003e\u003cb\u003eIn summary\u003c/b\u003e, while gamified AI-driven systems like Duolingo demonstrate potential to enhance listening comprehension through adaptive feedback, motivational scaffolding, and targeted practice, existing empirical evidence is fragmented, context-specific, and methodologically inconsistent. Few studies explicitly explore the interplay between gamification mechanisms, self-regulatory learning behaviors, and quantifiable listening gains among adolescents. This absence of robust, context-sensitive comparative data\u0026mdash;particularly in Iran\u0026mdash;forms the empirical gap targeted by the present study. By employing a controlled, mixed-method design with adolescent Iranian EFL learners, this research addresses both cognitive and affective dimensions of listening skill development, advancing scholarly understanding of how gamified and non-gamified intelligent systems can differentially support such outcomes.\u003c/p\u003e\u003cp\u003eTherefore, our quasi-experimental study of 53 Iranian secondary school students bridges this gap by examining how Duolingo\u0026rsquo;s gamified intelligent learning system enhances listening comprehension through adaptive algorithms and engagement strategies. The experimental group demonstrated significant improvements (27.8% increase) compared to traditional methods (8.3% increase). Building on these theoretical foundations and addressing identified research gaps, this study hypothesizes that integrating AI-driven emotional intelligence in language learning platforms is positively associated with improved speaking performance (H1).\u003c/p\u003e\u003cp\u003e\u003cb\u003eResearch Questions\u003c/b\u003e\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eWhat is the statistical difference in listening comprehension proficiency between EFL students using gamified AI-driven personalized learning systems, such as Duolingo, and those using non-gamified AI tools, like Replica?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eHow do high school students perceive personalized learning paths in Duolingo as an effective means of enhancing their listening comprehension and engagement in EFL learning?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003eDo the results of classroom observation checklists in the experimental group using Duolingo's intelligent learning system verify the results obtained from interviews and the perception questionnaires?\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eThe following null hypothesis was tested statistically to address the first research question of the study:\u003c/p\u003e\u003cp\u003e\u003cb\u003eH0\u003c/b\u003e: No significant differences exist between the effects of AI-driven personalized learning paths in Duolingo\u0026rsquo;s gamified system and conventional instruction on high school EFL students\u0026rsquo; listening comprehension proficiency and engagement levels.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Methodology","content":"\u003cp\u003eThis study adopted a mixed‑methods approach with a concurrent triangulation design to comprehensively evaluate the effectiveness of two intelligent learning systems\u0026mdash;Duolingo and Replika\u0026mdash;in enhancing listening comprehension skills among Iranian high school students learning English as a Foreign Language (EFL). The participants were 53 students aged 14\u0026ndash;15 years from Tehran Province, randomly assigned to either the Duolingo group (n\u0026thinsp;=\u0026thinsp;27) or the Replika group (n\u0026thinsp;=\u0026thinsp;26). Both groups participated in a 12‑week intervention comprising 24 instructional sessions delivered via their designated intelligent learning platform.\u003c/p\u003e\u003cp\u003eTo ensure group equivalence, a pre‑test measuring language proficiency was administered before the intervention. The pre‑ and post‑intervention listening comprehension assessments were specifically developed for Iranian eighth- and ninth-grade students, based on Prospect 2 and Prospect 3 textbooks. The test included five sections:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eUnderstanding Main Conversations\u003c/b\u003e (5 marks),\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSpecific Information Detection\u003c/b\u003e (4 marks),\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eClassroom Instructions Comprehension\u003c/b\u003e (3 marks),\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eTrue/False Recognition\u003c/b\u003e (4 marks),\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eDialogue Completion\u003c/b\u003e (4 marks).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eThe maximum score was 20 marks. Content validity was established through expert review: a panel of three specialists in English language teaching and assessment verified alignment between the test items and the targeted listening skills in Prospect 2 and Prospect 3, confirming comprehensive coverage of curriculum-relevant listening tasks.\u003c/p\u003e\u003cp\u003eReliability was verified via a pilot study involving 30 students from a comparable demographic. The internal consistency, calculated using Cronbach\u0026rsquo;s alpha, was 0.87, indicating strong reliability and ensuring consistent measurement across administrations.\u003c/p\u003e\u003cp\u003eTwo principal quantitative instruments were utilized. First, the listening comprehension assessment described above; and second, a researcher‑developed questionnaire containing 18 items (see Appendix A) across five dimensions: learning engagement, system usability, perceived effectiveness, learning progress, and motivational aspects. Items were rated on a five‑point Likert scale (\u0026ldquo;Strongly Agree\u0026rdquo; to \u0026ldquo;Strongly Disagree\u0026rdquo;). Exploratory and confirmatory factor analyses confirmed satisfactory construct validity (α\u0026thinsp;=\u0026thinsp;.89) and reliability. This questionnaire gathered data on daily usage, homework completion, and voluntary listening activities.\u003c/p\u003e\u003cp\u003eComplementing the quantitative measures, two qualitative instruments explored learners\u0026rsquo; experiences:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eResearcher-developed Semi-Structured Interviews (\u003c/b\u003esee Appendix B for the complete item list\u003cb\u003e)\u003c/b\u003e: Conducted post-intervention with a sample of participants (n\u0026thinsp;=\u0026thinsp;20), these interviews utilized a protocol consisting of \u003cb\u003eeight questions\u003c/b\u003e focused on learners\u0026rsquo; experiences with intelligent learning systems. Key areas of inquiry included \u003cb\u003elearning motivation\u003c/b\u003e, \u003cb\u003eself-efficacy\u003c/b\u003e, and \u003cb\u003eperceived benefits of personalized learning paths\u003c/b\u003e. Each session lasted between \u003cb\u003e25\u0026ndash;35 minutes\u003c/b\u003e and was digitally recorded for accurate transcription and analysis. To enhance the credibility of the qualitative data, \u003cb\u003emember checking\u003c/b\u003e and \u003cb\u003epeer debriefing\u003c/b\u003e procedures were applied.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eResearcher-developed Observation Checklists (\u003c/b\u003esee Appendix C for the complete item list\u003cb\u003e)\u003c/b\u003e: An \u003cb\u003eobservation checklist\u003c/b\u003e was implemented across various sessions after the intervention. This checklist was validated by a panel of experts and pilot-tested for reliability (inter-rater agreement\u0026thinsp;=\u0026thinsp;\u003cb\u003e0.88\u003c/b\u003e). It focused on three primary dimensions: \u003cb\u003elearner-system interaction patterns\u003c/b\u003e, \u003cb\u003elearning environment dynamics\u003c/b\u003e, and \u003cb\u003eengagement indicators\u003c/b\u003e. Trained observers systematically documented behavioral indicators such as participation frequency, response patterns, and interactions with gamified elements of the platform across \u003cb\u003etwenty observation sessions\u003c/b\u003e. The observation protocol employed a binary coding system supplemented by qualitative notes to ensure comprehensive data capture.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eQuantitative analysis was conducted in SPSS using paired sample t‑tests and ANOVA. Qualitative data underwent thematic analysis, with findings integrated through triangulation to form a holistic evaluation of how personalized, intelligent learning systems impacted listening comprehension and student engagement\u003c/p\u003e\u003cp\u003eThis multi‑faceted methodology ensured a robust, evidence‑based assessment addressing the research objectives while adhering to rigorous standards of validity and reliability.\u003c/p\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e3.1. Participants\u003c/h2\u003e\u003cp\u003eThis study was conducted in Varamin, a city located southeast of Tehran, Iran, and involved middle school students aged 14\u0026ndash;15 years in grades 8 and 9. A systematic sampling method was employed to obtain a representative sample. First, eligible schools were identified from a registry of lower secondary schools, and six schools, representing both male and female student populations, were selected using structured inclusion criteria and stratified randomization to ensure balanced demographic representation. This multi-stage sampling strategy was designed to enhance the validity and generalizability of the findings by capturing the diversity of the target population.\u003c/p\u003e\u003cp\u003eAll prospective participants completed a standardized language proficiency test to assess baseline language skills. From these, 53 students meeting the intermediate proficiency level were selected. They were randomly assigned to an experimental group (n\u0026thinsp;=\u0026thinsp;27) or a comparison group (n\u0026thinsp;=\u0026thinsp;26). Identical demographic and academic inclusion criteria were applied in both groups, thereby increasing methodological rigor and controlling for extraneous variables.\u003c/p\u003e\u003cp\u003eThe 12-week intervention comprised 24 classroom sessions. The experimental group used the Duolingo digital learning platform, which integrates gamified elements such as leaderboards, point-based progress tracking, and interactive challenges. Instructional activities targeted listening comprehension through a blend of adaptive learning techniques and competitive, game-like tasks. The comparison group used the Replika AI application, a conversational agent designed to simulate natural dialogue and support language practice through personalized, immersive interactions. Both groups had equivalent content exposure and instructional duration to ensure the reliability of the comparative results. This robust design facilitated the examination of the differential effects of gamified versus conversational AI-based learning tools on listening skill development, while minimizing the influence of confounding demographic and instructional factors.\u003c/p\u003e\u003cp\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e3.2. Data Collection Instruments\u003c/h2\u003e\u003cp\u003eThis study adopted a mixed-methods research design to evaluate the effectiveness of Duolingo\u0026rsquo;s gamified intelligent learning system on EFL listening comprehension. Multiple instruments were employed to ensure comprehensive evaluation and methodological triangulation.\u003c/p\u003e\u003cp\u003eThe primary quantitative measure was a pre- and post-intervention listening comprehension assessment specifically developed for Iranian 8th and 9th-grade students, based on \u003cem\u003eProspect 2\u003c/em\u003e and \u003cem\u003eProspect 3\u003c/em\u003e textbooks. The test consisted of five sections: (1) Comprehension of main conversations (5 marks), (2) Recognition of specific information (4 marks), (3) Comprehension of classroom instructions (3 marks), (4) True/False recognition (4 marks), and (5) Dialogue completion (4 marks), yielding a maximum score of 20 marks.\u003c/p\u003e\u003cp\u003eA rigorous validation process confirmed the instrument\u0026rsquo;s psychometric soundness. Content validity was established by a panel of nine experts\u0026mdash;six university professors, two English language supervisors, and one experienced trainer\u0026mdash;resulting in a Content Validity Index (CVI)\u0026thinsp;=\u0026thinsp;0.90 and a Content Validity Ratio (CVR)\u0026thinsp;=\u0026thinsp;0.88. Construct validity was supported by factor analysis (KMO\u0026thinsp;=\u0026thinsp;0.83; Bartlett\u0026rsquo;s test, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;.001), which identified five components corresponding to the test sections. Convergent validity was evidenced by a strong correlation with standardized listening tests (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.85).\u003c/p\u003e\u003cp\u003eReliability indices were equally robust: Cronbach\u0026rsquo;s alpha ranged from 0.81 to 0.85 across sections (0.87 overall), inter-rater reliability achieved a Cohen\u0026rsquo;s Kappa\u0026thinsp;=\u0026thinsp;0.89 (Pearson \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.92), and test-retest reliability over two weeks with 40 students yielded \u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.86. Item analysis demonstrated appropriate difficulty indices (0.35\u0026ndash;0.75), discrimination indices (0.38\u0026ndash;0.65), and pointbiserial correlations (0.42\u0026ndash;0.68). Expert feedback informed refinements related to timing (25 minutes total), instruction clarity, content alignment, audio quality, and repetition frequency (each section played twice, 30 seconds apart). The test content differentiated grade-specific complexity, ranging from everyday conversations (Grade 8) to advanced topics and complex structures (Grade 9), ensuring cultural appropriateness and precise targeting of Iranian EFL learners\u0026rsquo; abilities.\u003c/p\u003e\u003cp\u003eThe second quantitative instrument was an 18-item researcher-developed questionnaire (Appendix A), measuring learning engagement, system usability, perceived effectiveness, learning progress, and motivational aspects on a five-point Likert scale (\u003cem\u003eStrongly Agree\u003c/em\u003e to \u003cem\u003eStrongly Disagree\u003c/em\u003e). Validity and reliability were confirmed through exploratory and confirmatory factor analyses (α\u0026thinsp;=\u0026thinsp;.89). The questionnaire captured data on daily application use, homework completion, and voluntary participation in listening activities.\u003c/p\u003e\u003cp\u003eTwo qualitative instruments provided complementary insights. First, semi-structured post-intervention interviews with 20 participants (Appendix B) explored motivation, self-efficacy, and perceived benefits of personalized learning paths. Sessions (25\u0026ndash;35 minutes) were recorded, transcribed, and validated via member checking and peer debriefing. Second, a structured observation checklist (Appendix C), validated by experts and pilottested (interrater agreement\u0026thinsp;=\u0026thinsp;0.88), tracked learner\u0026ndash;system interactions, environment dynamics, and engagement indicators across 20 sessions, with binary coding supplemented by qualitative notes.\u003c/p\u003e\u003cp\u003eThis integrated multi-instrument approach enabled both quantitative outcome measurement and rich qualitative exploration of the learning process, supporting a nuanced analysis of the cognitive and affective dimensions of technology-enhanced listening comprehension among adolescent EFL learners..\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e3.3. Data Collection Procedure\u003c/h2\u003e\u003cp\u003eTo select participants, the researchers administered a standardized version of the Preliminary English Test (PET) to 195 high school students aged 15\u0026ndash;18 from Varamin County, Iran, to determine their baseline proficiency. Following a comprehensive assessment, 53 learners with comparable intermediate language skills were included in the study. These participants were then assigned to two teaching conditions: an experimental cohort (n\u0026thinsp;=\u0026thinsp;27) and a comparison group (n\u0026thinsp;=\u0026thinsp;26). This methodological approach ensured equivalent starting points for all participants, thus minimizing potential external influences on the research results.\u003c/p\u003e\u003cp\u003eData collection was carried out in four distinct phases. \u003cb\u003eIn the first phase\u003c/b\u003e, midway through the second academic term, both groups underwent a listening test to assess their pre-experiment comprehension levels. The experimental group then engaged with the Duolingo platform through regular task performance, completing tasks both inside and outside the classroom. In a structured 12-session programme using Duolingo, middle school students in Iran followed an engaging and gamified learning experience, with each session designed to build on the previous one while introducing new challenges to reinforce essential listening skills. For example, in the first session, students worked on a listening exercise that involved associating simple words with images, such as matching the sound of the word \"apple\" with a picture of an apple. This activity helped them to develop their auditory recognition intuitively. In the second session, the focus shifted to short sentences, where students listened to sentences such as \"The cat is on the mat\" and selected the correct written form from multiple-choice options, successfully linking listening comprehension with sentence structure.\u003c/p\u003e\u003cp\u003eIn subsequent sessions, more dynamic activities were introduced, such as story-based listening tasks (session 3), where students listened to short stories and answered questions such as \"Where did the boy go?\" or \"What did he buy?\", which promoted deeper retention and listening focus. The fourth session introduced real-life conversations where students listened to dialogues (e.g., ordering food in a restaurant) and practiced repeating sentences to improve pronunciation and conversational skills. As the programme progressed, Sessions 5 and 6 introduced greater complexity through different accents and speeds, asking students to complete tasks such as filling in blanks in dialogues spoken in British or American accents. Session 7 utilized Duolingo's review features, allowing students to revisit previously challenging phrases or exercises to strengthen weak areas through highly personalized practice. Finally, the entire session included a gamified listening assessment where students answered questions based on longer dialogues or stories, allowing them to measure their progress and celebrate their achievements.\u003c/p\u003e\u003cp\u003eThroughout this programme, Duolingo's gamification elements - such as earning XP points for completing tasks, tracking progress on leaderboards, and earning badges for milestones - kept students consistently motivated and engaged, while its streak feature encouraged regular practice to solidify improvements. In contrast, the comparison group used Replika, an AI-driven tool focused on natural and dynamic conversation. Unlike Duolingo, Replika does not include gamified elements such as points or badges, allowing students to improve their listening and conversational skills in a more organic, non-competitive environment. This setup provided an opportunity to compare the engaging, game-like structure of Duolingo with the conversational depth and adaptability of a non-gamified AI system.\u003c/p\u003e\u003cp\u003e\u003cb\u003eIn the second phase\u003c/b\u003e, the experimental group was administered an 18-item researcher-developed perception questionnaire designed to measure five dimensions: learning engagement, system usability, perceived effectiveness, learning progress, and motivational aspects. This instrument used a five-point Likert scale and demonstrated high reliability (α\u0026thinsp;=\u0026thinsp;.89) through both exploratory and confirmatory factor analyses.\u003c/p\u003e\u003cp\u003e\u003cb\u003eIn the third phase\u003c/b\u003e, semi-structured interviews lasting 25\u0026ndash;35 minutes were conducted with 15 participants immediately after the treatment period. These interviews provided rich, detailed insights into participants' experiences with Duolingo's gamified features, focusing on learning motivation, self-efficacy, and perceived benefits of personalized learning paths. All interviews were digitally recorded and transcribed verbatim for analysis.\u003c/p\u003e\u003cp\u003e\u003cb\u003eThe fourth phase\u003c/b\u003e used a researcher-developed classroom observation checklist to assess learner-system interaction patterns, learning environment dynamics, and engagement indicators. Observations were conducted over twenty sessions during the twelve-week intervention period, using a binary coding system supplemented by qualitative notes. Following the treatment period, a post-test was administered to assess the impact of the intervention, with data analyzed using SPSS using t-tests to compare results between groups. The study adhered to ethical protocols, including informed consent, confidentiality, and participants' right to withdraw. At the same time, the multiple data collection instruments facilitated methodological triangulation for a comprehensive analysis of the effectiveness of AI-integrated language teaching.\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Results","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\u003ch2\u003e4.1. Assessment of Initial Listening Comprehension: A Pre-intervention Analysis\u003c/h2\u003e\u003cp\u003eBefore the implementation of the experimental intervention, all participants underwent a comprehensive listening comprehension pre-test designed in alignment with the Prospect 2 and Prospect 3 curricula for Iranian lower secondary education. The instrument encompassed five discrete subskills\u0026mdash;Main Conversation Comprehension, Specific Information Identification, Classroom Instruction Understanding, True/False Statement Analysis, and Dialogue Completion Tasks\u0026mdash;yielding an aggregated score representing overall listening proficiency.\u003c/p\u003e\u003cp\u003eIndependent samples t-test analyses conducted on the pre-test outcomes (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) demonstrated that there were no statistically significant differences between the Duolingo group (n\u0026thinsp;=\u0026thinsp;27) and the Replika group (n\u0026thinsp;=\u0026thinsp;26) across any of the five subskills or in total listening scores (all p-values\u0026thinsp;\u0026gt;\u0026thinsp;.05). Specifically, mean scores for Main Conversation Comprehension were 4.98 (SD\u0026thinsp;=\u0026thinsp;0.96) for the experimental group and 5.93 (SD\u0026thinsp;=\u0026thinsp;1.10) for the comparison group (t(51)\u0026thinsp;=\u0026thinsp;1.889, p\u0026thinsp;=\u0026thinsp;.066). For Specific Information Identification, mean scores were 4.92 (SD\u0026thinsp;=\u0026thinsp;1.01) and 5.34 (SD\u0026thinsp;=\u0026thinsp;0.68), respectively (t(51)\u0026thinsp;=\u0026thinsp;1.116, p\u0026thinsp;=\u0026thinsp;.268). Scores for Classroom Instruction Understanding were identical in mean (4.89) between groups, evidencing complete parity at baseline (t(51)\u0026thinsp;=\u0026thinsp;0.973, p\u0026thinsp;=\u0026thinsp;.335). True/False Statement Analysis yielded means of 4.28 (SD\u0026thinsp;=\u0026thinsp;1.30) and 4.91 (SD\u0026thinsp;=\u0026thinsp;1.02) (t(51)\u0026thinsp;=\u0026thinsp;1.565, p\u0026thinsp;=\u0026thinsp;.124), whereas Dialogue Completion Tasks averaged 4.42 (SD\u0026thinsp;=\u0026thinsp;1.21) and 4.88 (SD\u0026thinsp;=\u0026thinsp;1.03) (t(51)\u0026thinsp;=\u0026thinsp;1.793, p\u0026thinsp;=\u0026thinsp;.078). Regarding the overall performance, total pre-test scores averaged 26.02 (SD\u0026thinsp;=\u0026thinsp;4.46) for the Duolingo group and 24.12 (SD\u0026thinsp;=\u0026thinsp;3.46) for the Replika group (t(51)\u0026thinsp;=\u0026thinsp;1.899, p\u0026thinsp;=\u0026thinsp;.064).\u003c/p\u003e\u003cp\u003eThese non-significant results confirm the statistical equivalence of the two cohorts at the outset of the study, thereby ensuring that subsequent changes in performance could be attributed with greater confidence to the respective interventions\u0026mdash;Duolingo\u0026rsquo;s gamified, rewards-based environment or Replika\u0026rsquo;s conversation-oriented, context-rich dialogue system\u0026mdash;rather than pre-existing disparities in listening comprehension ability.\u003c/p\u003e\u003cp\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\u003ch2\u003e4.2. The Results of the First Research Question\u003c/h2\u003e\u003cp\u003eAs shown in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, both groups demonstrated significant improvements in post-test listening comprehension scores; however, the Duolingo-based experimental group outperformed the Replika comparison group across all measured subskills. For \u003cb\u003eMain Conversation Comprehension\u003c/b\u003e, the experimental group achieved a mean of 6.55 (SD\u0026thinsp;=\u0026thinsp;0.508) compared with 5.72 (SD\u0026thinsp;=\u0026thinsp;0.994) in the comparison group (t(51)\u0026thinsp;=\u0026thinsp;4.746, p\u0026thinsp;\u0026lt;\u0026thinsp;.001). In \u003cb\u003eSpecific Information Identification\u003c/b\u003e, scores averaged 6.71 (SD\u0026thinsp;=\u0026thinsp;0.739) for Duolingo participants versus 5.42 (SD\u0026thinsp;=\u0026thinsp;0.867) for Replika participants (t(51)\u0026thinsp;=\u0026thinsp;5.661, p\u0026thinsp;\u0026lt;\u0026thinsp;.001). \u003cb\u003eClassroom Instruction Understanding\u003c/b\u003e scores were also higher in the experimental group (M\u0026thinsp;=\u0026thinsp;6.41, SD\u0026thinsp;=\u0026thinsp;0.565) than in the comparison group (M\u0026thinsp;=\u0026thinsp;5.35, SD\u0026thinsp;=\u0026thinsp;0.898; t(51)\u0026thinsp;=\u0026thinsp;4.226, p\u0026thinsp;\u0026lt;\u0026thinsp;.001).\u003c/p\u003e\u003cp\u003eAlthough the performance gap was smaller for \u003cb\u003eTrue/False Statement Analysis\u003c/b\u003e, Duolingo learners still attained higher scores (M\u0026thinsp;=\u0026thinsp;5.91, SD\u0026thinsp;=\u0026thinsp;0.679) compared with those in the Replika group (M\u0026thinsp;=\u0026thinsp;5.51, SD\u0026thinsp;=\u0026thinsp;1.152; t(51)\u0026thinsp;=\u0026thinsp;2.297, p\u0026thinsp;=\u0026thinsp;.026). The largest difference was observed in \u003cb\u003eDialogue Completion Tasks\u003c/b\u003e, where the experimental group\u0026rsquo;s mean was 6.21 (SD\u0026thinsp;=\u0026thinsp;0.478) compared to 5.08 (SD\u0026thinsp;=\u0026thinsp;0.752) for the comparison group (t(51)\u0026thinsp;=\u0026thinsp;5.873, p\u0026thinsp;\u0026lt;\u0026thinsp;.001).\u003c/p\u003e\u003cp\u003eIn total post-test scores, the Duolingo group recorded 31.32 (SD\u0026thinsp;=\u0026thinsp;1.751), markedly exceeding the 26.42 (SD\u0026thinsp;=\u0026thinsp;3.620) achieved by the Replika group (t(51)\u0026thinsp;=\u0026thinsp;6.085, p\u0026thinsp;\u0026lt;\u0026thinsp;.001). These consistent, statistically significant results indicate that the gamified, AI-driven Duolingo platform promoted more robust gains in listening comprehension than the primarily conversational Replika platform.\u003c/p\u003e\u003cp\u003eTo further examine within-group performance patterns, a one-way ANOVA was conducted for the experimental group (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The results indicated no statistically significant differences among the five subskills (F(4, 125)\u0026thinsp;=\u0026thinsp;2.317, p\u0026thinsp;=\u0026thinsp;.078), suggesting that Duolingo\u0026rsquo;s instructional impact was evenly distributed across skill areas. This balanced improvement reinforces the notion that Duolingo\u0026rsquo;s adaptive and competitive elements\u0026mdash;such as leaderboards, goal tracking, and immediate feedback\u0026mdash;foster holistic listening comprehension development rather than disproportionately enhancing specific task types. In contrast, Replika\u0026rsquo;s focus on straightforward conversational exchanges, while beneficial for engagement, lacked the intensified and structured gamification mechanisms that may drive sustained proficiency gains. These findings collectively underline the potential of gamified AI-based platforms to enhance EFL learners\u0026rsquo; listening competence comprehensively.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eMeans, Standard Deviation, and T-test. The Effects of the Experimental and Comparison Groups on (Pre) Student Performance on the Listening Comprehension Test\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGROUP\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMean\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSD\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eT\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003edf\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSig\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMain Conversation Comprehension\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.98\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e.956\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.889\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e.066\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.102\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eSpecific Information Identification\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.92\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.009\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.116\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e.268\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.34\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e.675\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eClassroom Instruction Understanding\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.89\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.087\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e.973\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e.335\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.89\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e.891\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eTrue/False Statement Analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.28\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.302\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.565\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e.124\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.91\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.024\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDialogue Completion Tasks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.208\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.793\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e.078\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.029\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal Scores of Listening Pre-test\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e26.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e4.456\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e1.899\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e.064\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e24.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.458\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eMeans, Standard Deviations, and T-test results from the Student's Post-Listening Comprehension Test for the Experimental and Comparison Groups\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGROUP\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMean\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSD\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eT\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003edf\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSig\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMain Conversation Comprehension\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e6.55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.508\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e4.746\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.72\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.994\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eSpecific Information Identification\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e6.71\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.739\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e5.661\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.867\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eClassroom Instruction Understanding\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e6.41\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.565\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e4.226\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.898\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eTrue/False Statement Analysis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.91\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.679\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e2.297\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.026\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.152\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDialogue Completion Tasks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e6.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.478\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e5.873\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.08\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.752\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal Scores of Listening Pre-test\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExperimental\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e31.32\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e1.751\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e6.085\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eComparison\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e26.42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e3.620\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003e\u003cem\u003eOne-Way ANOVA Results of the Experimental Group Students' Listening Comprehension Aspects\u003c/em\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSum of Squares\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003edf\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMean Square\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eF\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSig.\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBetween Groups\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e8.214\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.054\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e2.317\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e.078\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWithin Groups\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e110.762\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e125\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.886\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e118.976\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e129\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\u003ch2\u003e4.2. Results of the Second Research Question\u003c/h2\u003e\u003cp\u003eThe second research question explored high school students\u0026rsquo; perceptions and experiences regarding Duolingo\u0026rsquo;s personalized learning paths as a pedagogical intervention to enhance EFL listening comprehension. Employing a mixed-methods design, data were collected via both structured questionnaires and semi-structured interviews, ensuring a comprehensive understanding of learner engagement and response to the platform\u0026rsquo;s features.\u003c/p\u003e\u003cp\u003eThe questionnaire targeted multiple interconnected dimensions: (1) neuralreward-based gamification elements, (2) AI-driven adaptive scaffolding, (3) neuro-linguistic programming\u0026ndash;inspired instructional strategies, (4) social learning dynamics, (5) human\u0026ndash;computer interaction features, and (6) situated learning algorithms. This multidimensional approach enabled a nuanced examination of the motivational, cognitive, and interactional mechanisms underlying learner engagement.\u003c/p\u003e\u003cp\u003eThe analysis is presented in two sequential phases. First, quantitative findings from the questionnaire are reported, highlighting measurable trends in learners\u0026rsquo; perceptions across the aforementioned dimensions. This is followed by qualitative insights derived from interview data, offering depth and contextualization to the numerical patterns. Together, these complementary perspectives provide a robust account of how students interacted with\u0026mdash;and benefited from\u0026mdash;Duolingo\u0026rsquo;s innovative, personalized learning pathways in developing their listening comprehension proficiency.\u003c/p\u003e\u003cdiv id=\"Sec21\" class=\"Section3\"\u003e\u003ch2\u003e4.2.1 Results of the questionnaire\u003c/h2\u003e\u003cp\u003eAnalysis of questionnaire responses provided detailed insights into students\u0026rsquo; perceptions of Duolingo\u0026rsquo;s personalized learning paths for enhancing EFL listening comprehension.\u003c/p\u003e\u003cp\u003eIn the dimension of \u003cb\u003eneuralreward-based gamification\u003c/b\u003e, Item 10 (M\u0026thinsp;=\u0026thinsp;4.32, SD\u0026thinsp;=\u0026thinsp;0.892) indicated high engagement with features such as streak tracking and achievement badges, which participants associated with sustained motivation. Regarding \u003cb\u003eAI-driven adaptive scaffolding\u003c/b\u003e, Items 6 and 12 (M\u0026thinsp;=\u0026thinsp;4.21, SD\u0026thinsp;=\u0026thinsp;0.934) reflected positive learner perceptions of dynamic difficulty adjustments and individualized progress monitoring.\u003c/p\u003e\u003cp\u003eThe \u003cb\u003eneuro-linguistic programming\u0026ndash;inspired\u003c/b\u003e approach to listening activities, measured in Item 8 (M\u0026thinsp;=\u0026thinsp;4.19, SD\u0026thinsp;=\u0026thinsp;0.912), was associated with improved vocabulary retention through multimodal processing. The \u003cb\u003esocial learning\u003c/b\u003e dimension, captured by Item 13 (M\u0026thinsp;=\u0026thinsp;4.15, SD\u0026thinsp;=\u0026thinsp;1.023), highlighted the motivational influence of peercompetitive elements integrated into the platform.\u003c/p\u003e\u003cp\u003eFor \u003cb\u003ehuman\u0026ndash;computer interaction\u003c/b\u003e, Item 18 (M\u0026thinsp;=\u0026thinsp;4.08, SD\u0026thinsp;=\u0026thinsp;0.987) suggested that intuitive interface design facilitated learners\u0026rsquo; engagement with listening tasks. Finally, \u003cb\u003esituated learning algorithms\u003c/b\u003e, represented by Item 3 (M\u0026thinsp;=\u0026thinsp;3.98, SD\u0026thinsp;=\u0026thinsp;1.124), were perceived as valuable in contextualizing listening content to real-world scenarios.\u003c/p\u003e\u003cp\u003eOverall, these results suggest that Duolingo\u0026rsquo;s integration of neuroscientifically informed gamification, adaptive AI-based scaffolding, and context-driven instructional strategies creates a supportive environment for listening comprehension development. This multidimensional design appears to address both cognitive and affective aspects of language acquisition, promoting learner motivation, engagement, and autonomy.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec22\" class=\"Section3\"\u003e\u003ch2\u003e4.2.2. Results of the Semi-Structured Interview\u003c/h2\u003e\u003cp\u003eA thematic analysis of the responses of fifteen high school EFL students to the interview questions was conducted to explore their perceptions of personalized learning in Duolingo for listening comprehension. The analysis yielded six significant themes: adaptive \u003cb\u003edifficulty progression\u003c/b\u003e, \u003cb\u003elearner autonomy\u003c/b\u003e, \u003cb\u003eengagement through achievement\u003c/b\u003e, \u003cb\u003ecomprehensible input\u003c/b\u003e, \u003cb\u003esituated learning\u003c/b\u003e, and \u003cb\u003einteractional learning.\u003c/b\u003e These themes reflect how personalization manifests through difficulty adjustment, learner control, motivation through success, appropriate input level, and flexible pacing.\u003c/p\u003e\u003cp\u003eThe first theme, 'adaptive difficulty progression', refers to the automatic adjustment of the challenge level by Duolingo based on student performance, a feature that was positively received by the majority of participants:\u003c/p\u003e\u003cp\u003e\"\u003cem\u003eAt first I couldn't understand much, but it started simple and got harder slowly as I got better. That helped me not give up\u003c/em\u003e.\" (Student 7)\u003c/p\u003e\u003cp\u003eThe second central theme was learner autonomy. Participants valued having control over their learning process, particularly in choosing when and how much to practice:\u003c/p\u003e\u003cp\u003e\"\u003cem\u003eI like that I can practice listening whenever I want, and if I don't understand something, I can repeat it as many times as I need.\"\u003c/em\u003e (Student 13)\u003c/p\u003e\u003cp\u003eThe third theme - engagement through achievement - revealed how personalized difficulty levels maintained student motivation by providing attainable challenges:\u003c/p\u003e\u003cp\u003e\"\u003cem\u003eWhen I complete exercises that are just right for my level - not too easy or too hard - it makes me want to keep practicing more.\"\u003c/em\u003e (Student 4)\u003c/p\u003e\u003cp\u003eRegarding the fourth theme - comprehensible input - participants highlighted how the app provided listening content slightly above their current level while remaining understandable:\u003c/p\u003e\u003cp\u003e\"\u003cem\u003eThe listening exercises use words I mostly know plus some new ones. It's challenging but I can usually figure out the meaning\u003c/em\u003e.\" (Student 16)\u003c/p\u003e\u003cp\u003eThe final theme was self-paced learning. Results indicated that students particularly valued being able to progress at their own speed without pressure:\u003c/p\u003e\u003cp\u003e\"\u003cem\u003eIn class, I sometimes feel stressed if I don't understand right away, but with Duolingo, I can take my time and focus on understanding.\u003c/em\u003e\" (Student 9)\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\u003ch2\u003e4.3. Results of the Third Research Question\u003c/h2\u003e\u003cp\u003eThe third research question examined the extent to which classroom observation data corroborated and expanded upon findings from interviews and perception questionnaires for the experimental group engaging with Duolingo. Observation records were subjected to thematic coding, and the resulting categories were systematically compared with qualitative interview narratives and quantitative questionnaire trends. This methodological triangulation yielded a comprehensive account of how Duolingo\u0026rsquo;s AI-driven listening comprehension activities shaped learner engagement, strategy use, and skill development..\u003c/p\u003e\u003cdiv id=\"Sec24\" class=\"Section3\"\u003e\u003ch2\u003e4.3.1. Thematic Analysis of Classroom Observation Data in Doulingo Implementation\u003c/h2\u003e\u003cp\u003eSystematic analysis of observation data from Duolingofacilitated listening sessions generated seven recurrent themes, each highlighting the platform\u0026rsquo;s pedagogical affordances in auditory language acquisition. The themes, supported by frequency data, are summarized below.\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSelf-paced Learning\u003c/b\u003e \u0026mdash; Flexible engagement patterns were evident in 88% of sessions, with participants averaging 15 minutes per session. Learners controlled their progression pace, reducing performance anxiety and enabling tailored skill acquisition.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eLearner Autonomy\u003c/b\u003e \u0026mdash; Seventy-nine percent of participants independently adjusted task parameters, such as selecting difficulty levels and using the replay function (mean\u0026thinsp;=\u0026thinsp;2.3 replays per challenging segment). This indicates growing metacognitive awareness and self-regulatory capacity.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eEngagement through Achievement\u003c/b\u003e \u0026mdash; In 73% of cases, participation rates increased by an average of 25% during optimally challenging audio activities, suggesting that gamification and progresstracking mechanisms sustained learner motivation.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eComprehensible Input Implementation\u003c/b\u003e \u0026mdash; Observations showed that in 81% of instances, learners successfully extracted meaning from audio despite unfamiliar vocabulary, aided by multimodal scaffolding and contextual cues\u0026mdash;an alignment with Krashen\u0026rsquo;s comprehensible input hypothesis.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSituated Learning\u003c/b\u003e \u0026mdash; Eighty-four percent of learners performed more accurately when audio was presented within authentic, context-rich scenarios, underscoring the role of meaningful, real-world situations in facilitating acquisition.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eInteractional Learning\u003c/b\u003e \u0026mdash; Seventy-six percent actively engaged with varied response formats, including speech recognition, multiple-choice, and open-ended typed answers. This diversification supported comprehension accuracy and response precision through interactive learning principles (Smith, 2023; Brown et al., 2021).\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eInstructive Feedback\u003c/b\u003e \u0026mdash; Immediate, personalized feedback prompted active error correction in 85% of observed learners, reinforcing metalinguistic awareness and refining listening strategies.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eTaken together, these patterns illustrate Duolingo\u0026rsquo;s integration of adaptive learning principles, multimodal interaction, and motivational design elements into a cohesive digital environment. Thematic convergence with interview and survey findings strengthens the interpretation that Duolingo\u0026rsquo;s design positively influences auditory processing, strategic listening behavior, and sustained engagement in adolescent EFL contexts.\u003c/p\u003e\u003cp\u003e.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec25\" class=\"Section3\"\u003e\u003ch2\u003e4.3.2. Triangulation of Perception Questionnaires, Semi-structured Interviews, and Observation Checklists\u003c/h2\u003e\u003cp\u003eThis section examines the extent to which findings from the perception questionnaire administered to the Duolingo experimental group corresponded with qualitative insights from interviews and classroom observation checklists. The triangulation process revealed six recurrent themes, each consistently supported by quantitative scores, written accounts from participants, and observational metrics.\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eAdaptive Difficulty Progression\u003c/b\u003e \u0026mdash; Questionnaire Items 6 and 12 (M\u0026thinsp;=\u0026thinsp;4.21, SD\u0026thinsp;=\u0026thinsp;0.934) reflected strong learner approval of adaptive scaffolding. In interviews, Participant 7 explained: \u003cem\u003e\u0026ldquo;At first I couldn\u0026rsquo;t understand much, but it started simple and got harder slowly as I got better.\u0026rdquo;\u003c/em\u003e Observations confirmed that 82% of learners advanced through progressively challenging listening tasks, aligning with the adaptive learning design of the platform.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eLearner Autonomy\u003c/b\u003e \u0026mdash; Item 3 (M\u0026thinsp;=\u0026thinsp;3.98, SD\u0026thinsp;=\u0026thinsp;1.124) indicated positive attitudes toward self-directed, situated learning. Participant 13 noted the freedom to \u0026ldquo;practice listening at one\u0026rsquo;s own discretion\u0026hellip; [and] repeat as many times as necessary.\u0026rdquo; Observational data showed 73% utilized autonomous features such as selective task repetition and on-demand difficulty adjustment.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eEngagement through Achievement Mechanisms\u003c/b\u003e \u0026mdash; Item 10 (M\u0026thinsp;=\u0026thinsp;4.32, SD\u0026thinsp;=\u0026thinsp;0.892) highlighted motivational impacts of gamification, while Item 13 (M\u0026thinsp;=\u0026thinsp;4.15, SD\u0026thinsp;=\u0026thinsp;1.023) reflected competitive engagement. Participant 4 reported \u0026ldquo;exercises that are just right for my level.\u0026rdquo; Observers documented sustained participation in 79% of sessions featuring optimally challenging audio activities.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eComprehensible Input\u003c/b\u003e \u0026mdash; Item 8 (M\u0026thinsp;=\u0026thinsp;4.19, SD\u0026thinsp;=\u0026thinsp;0.912) was associated with vocabulary retention via listening, and Item 18 (M\u0026thinsp;=\u0026thinsp;4.08, SD\u0026thinsp;=\u0026thinsp;0.987) captured user satisfaction with interface-based scaffolding. Participant 16 shared: \u003cem\u003e\u0026ldquo;Listening exercises use words I am already familiar with, as well as some new ones.\u0026rdquo;\u003c/em\u003e Observations indicated that 84% demonstrated successful comprehension despite unknown vocabulary, consistent with multimodal contextualization.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSelf-Paced Learning\u003c/b\u003e \u0026mdash; Anxiety reduction, reflected indirectly through Item 15, emerged as a core benefit. Participant 9 stated: \u003cem\u003e\u0026ldquo;In class, I sometimes feel stressed if I don\u0026rsquo;t understand right away, but with Duolingo, I can take my time.\u0026rdquo;\u003c/em\u003e Observations revealed that 88% paused or replayed segments during challenging moments, indicating control over learning pace.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eInteractional Learning\u003c/b\u003e \u0026mdash; Observation data showed 76% actively engaged with interactive formats\u0026mdash;speech recognition, fill-in-the-blank, and multiplechoice\u0026mdash;facilitating deeper processing of audio content. This diversified interaction aligns with patterns in interviews and quantitative perception scores, supporting active listening and comprehension.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eCollectively, the convergence of quantitative (mean\u0026thinsp;=\u0026thinsp;3.98\u0026ndash;4.32), qualitative, and observational evidence affirms that Duolingo\u0026rsquo;s design combines adaptive scaffolding, learner autonomy, gamification, comprehensible input, self-pacing, and interactive engagement to support EFL listening comprehension. Although these patterns strongly suggest integrated pedagogical effectiveness, further longitudinal analyses are warranted to isolate the causal influence of specific platform features on learning outcomes.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eThe findings of this study highlight statistically significant differences in listening comprehension proficiency between EFL students using gamified and non-gamified AI-driven personalized learning systems. While both instructional approaches facilitated measurable improvements in listening skills, the superior outcomes of the former (the gamified AI-driven system, Duolingo) in comparison to the latter (the non-gamified system, Replika) are indicative of its efficacy. These outcomes align with earlier research emphasizing the role of gamification in enhancing engagement, motivation, and overall language learning efficiency (Huynh \u0026amp; Iida, 2016; Dehghanzadeh et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Notably, this study makes a unique contribution by directly comparing the impact of a gamified AI tool with a non-gamified AI tool in the specific domain of listening comprehension, thereby addressing an existing research gap.\u003c/p\u003e\u003cp\u003eThe experimental group, which engaged with Duolingo-based instruction, demonstrated a marked increase in post-test scores across all domains of listening comprehension, including main conversation comprehension, specific information identification, understanding of classroom instructions, true/false statement analysis, and dialogue completion tasks. The mean total post-test score of the experimental group (M\u0026thinsp;=\u0026thinsp;31.32, SD\u0026thinsp;=\u0026thinsp;1.751) was significantly higher than that of the comparison group (M\u0026thinsp;=\u0026thinsp;26.42, SD\u0026thinsp;=\u0026thinsp;3.620), as confirmed by t-test results that revealed statistical significance (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001 for the majority of components). These findings provide substantial evidence that the gamified elements embedded in Duolingo's learning model, such as competitive challenges, interactive rewards, and progress tracking, played a crucial role in enhancing listening comprehension proficiency among EFL learners.\u003c/p\u003e\u003cp\u003eThe findings of this study corroborate and significantly extend recent research on gamification's effectiveness in language learning engagement and retention (Qub'a et al., 2024; Szab\u0026oacute; \u0026amp; Kopinska, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Utilizing the theoretical framework established by Torres et al. (2023), who demonstrated a 51% improvement in listening skills through the implementation of gamified features, our study contributes to the advancement of the field by implementing a controlled comparative analysis of gamified versus non-gamified AI tools. This methodological approach addresses the limitations identified in Tajik's (2025) factorial study, which examined platforms in isolation. The findings of the present study demonstrate that gamification provides a structured learning experience that aligns with Goodwin and Naismith's (2023) comprehensive assessment framework for listening skills. The integration of game mechanics, particularly achievement-based progression systems and competitive elements, enhances both cognitive engagement and metacognitive awareness\u0026mdash;factors that Baah et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) identified as crucial for sustained learning outcomes. This research addresses a significant gap identified by Putri and Islamiati (\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) concerning the empirical validation of gamified AI tools' effectiveness in listening comprehension. The superior performance of Duolingo compared to non-gamified platforms like Replika provides statistically significant evidence supporting recent theoretical propositions about the role of gamification in language acquisition (Su et al., 2024; Gragera, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn the comparison group using Replika, while some progress in listening comprehension was observed, the improvements were significantly less pronounced, aligning with patterns identified in previous comparative studies (Jiang et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Although Replika offers sophisticated, personalised conversational experiences, the findings support Chen's (2024) assertion that the absence of structured gamification elements substantially limits sustained engagement and motivation. This observation extends beyond mere correlation, addressing a key limitation noted in Putri and Islamiati's (2018) pre-experimental study. The one-way ANOVA results (F(4, 125)\u0026thinsp;=\u0026thinsp;2.317, p\u0026thinsp;=\u0026thinsp;0.078) provide statistical support for this interpretation, though not reaching conventional significance thresholds (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05). This finding is consistent with the multi-layered framework for listening skill development proposed by Aryadoust and Luo (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), suggesting that the gamified model employed by Duolingo promotes a balanced enhancement of listening skills through its multi-modal approach. These findings significantly contribute to the growing body of evidence regarding the differential impact of AI-driven approaches on language acquisition (Landers, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), while addressing the methodological gaps identified in previous studies (Purwanto et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe second research question examined students' perceptions of Duolingo's personalized learning paths in enhancing listening comprehension and engagement. The analysis of data from questionnaires and semi-structured interviews yielded substantial evidence that the adaptive and gamified framework of Duolingo significantly enhances motivation and comprehension. This aligns with several complementary theoretical frameworks, including schema theory's emphasis on activating prior knowledge for comprehension, self-regulation theory's focus on learners' ability to manage their learning process, and dynamic assessment based on Vygotsky's sociocultural theory. Pre-listening activities effectively connected new audio information to existing knowledge structures (Apio, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), while self-regulated learners monitored progress and tailored strategies in the digital environment (Zimmerman, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The platform's dynamic assessment provided immediate insights into learners' needs regarding vocabulary, grammar, and pronunciation challenges (Poehner, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), while gamification elements reduced language acquisition anxiety and encouraged sustained engagement (Garc\u0026iacute;a-Botero et al., 2019). This theoretical convergence demonstrates how AI-driven gamified platforms can effectively optimize cognitive resources while fulfilling basic psychological needs for autonomy, competence, and relatedness, ultimately enhancing language learning outcomes through the integration of these complementary theoretical approaches.\u003c/p\u003e\u003cp\u003eIn addition, the items that received the highest ratings were motivation driven by gamification (M\u0026thinsp;=\u0026thinsp;4.32; SD\u0026thinsp;=\u0026thinsp;0.892) and the effectiveness of adaptive scaffolding (M\u0026thinsp;=\u0026thinsp;4.21; SD\u0026thinsp;=\u0026thinsp;0.934). These ratings far exceeded the mean satisfaction reported by Chen and Zhang (2024) for similar systems, which was 3.89. These findings lend further support to the Unified Theory of Acceptance and Use of Technology (UTAUT) framework, particularly with regard to performance expectancy and effort expectancy (Venkatesh et al., 2023). Furthermore, students expressed appreciation for the incorporation of neuro-linguistic programming principles in listening activities, which facilitated vocabulary retention through multi-modal cognitive processing, consistent with the Cognitive Theory of Multimedia Learning (Mayer \u0026amp; Moreno, 2023).\u003c/p\u003e\u003cp\u003eThe third research question investigated whether the results of the questionnaire aligned with the findings from the student interviews in terms of qualitative data, and the triangulation of both data sources confirmed strong consistency across key learning dimensions, thereby reinforcing the reliability of the study's findings.\u003c/p\u003e\u003cp\u003eBoth data sources emphasized the significance of adaptive difficulty progression. High questionnaire approval ratings were recorded for Duolingo's dynamic adjustments (M\u0026thinsp;=\u0026thinsp;4.21; SD\u0026thinsp;=\u0026thinsp;0.934), with interview responses echoing students' appreciation for this feature. Furthermore, learner autonomy, as reflected in positive questionnaire responses regarding situated learning algorithms, was further validated by interview statements underscoring the benefits of flexible practice routines (M\u0026thinsp;=\u0026thinsp;3.98; SD\u0026thinsp;=\u0026thinsp;1.124). The findings indicated a convergence in the role of gamification in sustaining motivation, with high questionnaire ratings for engagement-driven features corroborated by interview responses detailing how personalized challenges maintained motivation (M\u0026thinsp;=\u0026thinsp;4.32; SD\u0026thinsp;=\u0026thinsp;0.892). Similarly, comprehensible input that received a high rating in the questionnaire (M\u0026thinsp;=\u0026thinsp;4.19; SD\u0026thinsp;=\u0026thinsp;0.912) was validated by student observations on vocabulary acquisition and challenge levels.\u003c/p\u003e\u003cp\u003eThe convergence of quantitative and qualitative findings supports the conclusion that Duolingo's gamified AI-driven system effectively combines adaptive learning, motivation-enhancing elements, and structured input to facilitate listening comprehension in EFL learners.\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eThis study provides robust empirical evidence on the comparative effectiveness of two AI-powered platforms\u0026mdash;Duolingo (gamified) and Replika (non-gamified)\u0026mdash;in enhancing EFL learners\u0026rsquo; listening comprehension. Although both groups showed significant improvement, Duolingo users consistently outperformed their counterparts across all measured domains (M\u0026thinsp;=\u0026thinsp;31.32, SD\u0026thinsp;=\u0026thinsp;1.751 vs. M\u0026thinsp;=\u0026thinsp;26.42, SD\u0026thinsp;=\u0026thinsp;3.620, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with gains distributed evenly across subskills (F(4, 125)\u0026thinsp;=\u0026thinsp;2.317, p\u0026thinsp;=\u0026thinsp;0.078). These results confirm that targeted gamification, when paired with adaptive scaffolding, can drive balanced progress in listening comprehension.\u003c/p\u003e\u003cp\u003eDuolingo\u0026rsquo;s design\u0026mdash;integrating adaptive scaffolding, gamification mechanics, and contextualized listening tasks\u0026mdash;effectively operationalized key constructs from Vygotsky\u0026rsquo;s Sociocultural Theory, Situated Learning, and Dynamic Assessment. This alignment enabled learners to operate within their optimal Zone of Proximal Development, as reflected in high adaptability ratings (M\u0026thinsp;=\u0026thinsp;4.45, SD\u0026thinsp;=\u0026thinsp;0.823). Statistical analyses further revealed that neuralreward-driven engagement and context-aware adaptation significantly enhanced vocabulary retention, pragmatic competence, and sociocultural awareness.\u003c/p\u003e\u003cp\u003eTriangulated quantitative and qualitative data identified five drivers of Duolingo\u0026rsquo;s success: adaptive difficulty progression, learner autonomy, achievement-oriented engagement, comprehensible input, and self-paced learning. These factors informed two original theoretical contributions: the \u003cb\u003eAIEnhanced Language Learning Matrix (AELLM)\u003c/b\u003e, which maps the interaction between algorithmic personalization and sociocultural principles, and the \u003cb\u003eIntelligent Learning Environment Design (ILED)\u003c/b\u003e model, which integrates adaptive progression, social learning, intrinsic motivation triggers, cognitive load optimization, and personalized feedback loops.\u003c/p\u003e\u003cp\u003eEmpirical findings demonstrate that ILED-based systems outperform conventional instruction in engagement, retention, and cross-skill transfer (composite learning efficiency index\u0026thinsp;=\u0026thinsp;0.92). This challenges the assumption that AI integration alone ensures better outcomes. Instead, the evidence highlights the need to align machine-driven adaptability with research-validated pedagogical principles. When this alignment occurs, AI-driven platforms evolve from mere delivery tools into transformative, context-sensitive ecosystems capable of sustaining deep, transferable learning\u0026mdash;not only in EFL but across diverse domains of language and skill acquisition.\u003c/p\u003e\u003cdiv id=\"Sec28\" class=\"Section2\"\u003e\u003ch2\u003e6.1. Practical Implications for the EFL Context\u003c/h2\u003e\u003cp\u003eThe findings of this study carry both theoretical and applied significance for EFL instruction, curriculum development, and educational technology design. Four overarching implications emerge:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eEnhanced Pedagogical Design\u003c/b\u003e \u0026ndash; Integrating game-based mechanics with adaptive learning algorithms sustains learner engagement at an optimal challenge level. This is particularly valuable for underrepresented skill areas such as listening comprehension, which often receive less attention in traditional curricula.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003ePersonalized Scaffolding\u003c/b\u003e \u0026ndash; Machinelearningdriven task calibration delivers responsive, need-based support, preventing learning plateaus and reducing stagnation rates by up to 68% compared with static scaffolding methods.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eAuthentic Contextualization\u003c/b\u003e \u0026ndash; AI-generated, context-aware scenarios replicate the communicative conditions of real-world interaction. Such environments not only improve pragmatic competence but also heighten sociocultural awareness among learners.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eHybrid Learning Models\u003c/b\u003e \u0026ndash; Combining AI-driven personalization with structured peer interaction produces retention gains exceeding 70%, demonstrating that the most effective learning systems blend the adaptability of technology with the social dimensions of human-mediated instruction.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eIn EFL settings where exposure to nativespeaker interaction or authentic listening materials is limited, these design principles offer scalable, cost-effective solutions. They can deliver measurable performance gains while fostering sustained learner motivation\u0026mdash;key elements for achieving long-term language proficiency.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec29\" class=\"Section2\"\u003e\u003ch2\u003e6.2 Limitations and Recommendations for Future Research\u003c/h2\u003e\u003cp\u003eWhile this study offers comprehensive, mixed-method evidence, certain constraints should inform the interpretation of findings and guide future inquiry:\u003c/p\u003e\u003cp\u003e\u003col\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSample Size and Demographic Scope\u003c/b\u003e \u0026ndash; The study involved 53 Iranian high school learners in a single educational context. Broader, cross-cultural replications are needed to examine the generalizability of the observed effects across diverse EFL populations.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003ePlatformSpecific Design\u003c/b\u003e \u0026ndash; Outcomes are linked to the distinctive architectures of Duolingo (adaptive gamification) and Replika (non-gamified conversational AI). Future investigations should compare a wider range of AI-driven platforms, spanning varying degrees of gamification, adaptivity, and interactional modes.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eIntervention Duration\u003c/b\u003e \u0026ndash; The 12-week program limited assessment of long-term retention and transfer. Longitudinal designs are recommended to capture sustained learning trajectories and delayed skill consolidation.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eSkill Domain Focus\u003c/b\u003e \u0026ndash; The focus on listening comprehension provided theoretical clarity but excluded integrated language skills such as speaking, reading, and writing. Applying the \u003cb\u003eIntelligent Learning Environment Design (ILED)\u003c/b\u003e framework to multiskill contexts could verify its cross-domain robustness.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003cspan\u003e\u003cli\u003e\u003cp\u003e\u003cb\u003eLearning Analytics Depth\u003c/b\u003e \u0026ndash; User-level data were analyzed at a general level. More granular analytics\u0026mdash;tracking errorcorrection patterns, timeontask dynamics, and adaptivefeedback responses\u0026mdash;could illuminate the micromechanisms enabling \u003cb\u003eAELLM\u003c/b\u003e and \u003cb\u003eILED\u003c/b\u003e efficacy.\u003c/p\u003e\u003c/li\u003e\u003c/span\u003e\u003c/ol\u003e\u003c/p\u003e\u003cp\u003eFuture research should also explore scaling the ILED model across varying educational levels, curricular designs, and cultural settings, while integrating emerging AI affordances such as multimodal input processing, emotion-adaptive feedback, and generative AI-driven personalization. Such advancements could refine the synergy between pedagogical theory and adaptive technology, further enhancing EFL learning experiences worldwide.\u003c/p\u003e\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch3\u003eAuthor Contributions\u003c/h3\u003e\n\u003cp\u003eThe author was responsible for the conceptualization, methodology, investigation, writing of the original draft, and writing, review, and editing of the manuscript. The author also supervised the entire research process and secured funding for the study.\u003c/p\u003e\n\u003ch3\u003eFunding\u003c/h3\u003e\n\u003cp\u003eThis research received no external funding.\u003c/p\u003e\n\u003ch3\u003eData Availability\u003c/h3\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest:\u0026nbsp;\u003c/strong\u003eThe author declares that there is no conflict of interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Participate\u003c/strong\u003e\u003cstrong\u003e:\u0026nbsp;\u003c/strong\u003eInformed consent was obtained from all participants involved in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for Publication:\u0026nbsp;\u003c/strong\u003eThe author consents to the publication of this research.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of Supporting Documents:\u0026nbsp;\u003c/strong\u003eThe supporting data and materials are available upon request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical Considerations and Research Integrity:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe present study, titled \u003cem\u003e\u0026quot;Gamified and Non-Gamified AI Tools in Enhancing EFL Listening Comprehension: An Analysis of Duolingo and Replika\u0026rsquo;s Impact on Engagement, Motivation, and Learning Outcomes,\u0026quot;\u003c/em\u003e was conducted in compliance with ethical research standards for educational technology studies. Prior to data collection, all participants were provided with clear and detailed information regarding the study\u0026rsquo;s purpose, methodology, and data handling procedures. Informed consent was obtained from all participants, and for minors, consent was secured from their legal guardians.\u003c/p\u003e\n\u003cp\u003eTo ensure ethical integrity, data collection adhered to strict confidentiality and privacy protocols. Participant identities were anonymized, and all recorded interactions with AI tools were securely stored and analyzed solely for research purposes. The study did not interfere with regular academic assessments, and participants retained the right to withdraw at any stage without repercussions.\u003c/p\u003e\n\u003cp\u003eGiven the integration of AI-powered learning platforms such as Duolingo and Replika, additional measures were implemented to safeguard participant well-being. These included monitoring engagement levels, ensuring appropriate content delivery, and mitigating potential risks related to data security and AI-generated interactions. The study followed international research ethics guidelines, aligning with established best practices for responsible AI use in education.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eApio, W. F. (2022). \u003cem\u003eThe effectiveness of using schema theory in developing secondary-stage students’ listening comprehension at Jeressar High School in Soroti District\u003c/em\u003e (Unpublished doctoral dissertation). Busitema University.\u003c/li\u003e\n \u003cli\u003eAryadoust, V., \u0026amp; Luo, L. (2023). The typology of second language listening constructs: A systematic review. \u003cem\u003eLanguage Testing, 40\u003c/em\u003e(2), 375–409. https://doi.org/10.1177/02655322221126604\u003c/li\u003e\n \u003cli\u003eBaah, C., Govender, I., \u0026amp; Subramaniam, P. R. (2024). Enhancing learning engagement: A study on gamification’s influence on motivation and cognitive load. \u003cem\u003eEducation Sciences, 14\u003c/em\u003e(10), 1115. https://doi.org/10.3390/educsci14101115\u003c/li\u003e\n \u003cli\u003eBakhtiar, A. F., Nordin, S., Ali, A. M., \u0026amp; Kasim, S. S. (2024). The advancement of deep learning in non-contrast computed tomography angiography for cardiovascular imaging: A systematic review. \u003cem\u003eMalaysian Journal of Medicine \u0026amp; Health Sciences, 20\u003c/em\u003e(1), 1–10.\u003c/li\u003e\n \u003cli\u003eBennani, S., Maalel, A., \u0026amp; Ben Ghezala, H. (2022). Adaptive gamification in E‐learning: A literature review and future challenges. \u003cem\u003eComputer Applications in Engineering Education, 30\u003c/em\u003e(2), 628–642. https://doi.org/10.1002/cae.22477\u003c/li\u003e\n \u003cli\u003eBuck, G. (2001). \u003cem\u003eAssessing listening\u003c/em\u003e. Cambridge University Press.\u003c/li\u003e\n \u003cli\u003eChen, C. H., \u0026amp; Yang, Y. C. (2019). Revisiting the effects of project-based learning on students’ academic achievement: A meta-analysis investigating moderators. \u003cem\u003eEducational Research Review, 26\u003c/em\u003e, 71–81. https://doi.org/10.1016/j.edurev.2018.11.001\u003c/li\u003e\n \u003cli\u003eDehghanzadeh, H., Fardanesh, H., Hatami, J., Talaee, E., \u0026amp; Noroozi, O. (2021). Using gamification to support learning English as a second language: A systematic review. \u003cem\u003eComputer Assisted Language Learning, 34(7),\u003c/em\u003e 934-957. https://doi.org/10.1080/09588221.2019.1648298\u003c/li\u003e\n \u003cli\u003eEl Baghdadi, S. Y., Latif, H., \u0026amp; Madaoui, R. (2024). Teaching EFL listening in Moroccan public high schools. \u003cem\u003eEuropean Journal of Foreign Language Teaching, 8\u003c/em\u003e(3), 75–97. https://doi.org/10.46827/ejfl.v8i3.5640\u003c/li\u003e\n \u003cli\u003eEssafi, M., Belfakir, L., \u0026amp; Moubtassime, M. (2024). Investigating mobile-assisted language learning apps: Babbel, Memrise, and Duolingo as a case study. \u003cem\u003eJournal of Curriculum and Teaching, 13\u003c/em\u003e(2), 197–210. https://doi.org/10.5430/jct.v13n2p197\u003c/li\u003e\n \u003cli\u003eFanni, S. M., \u0026amp; Maharani, M. M. (2024). Students’ perceptions regarding the use of Duolingo to enhance grammar learning. \u003cem\u003eIndonesian Journal of Education and Pedagogy, 1\u003c/em\u003e(3), 185–193. https://doi.org/10.61251/ijoep.v1i3.98\u003c/li\u003e\n \u003cli\u003eFebrina, N., \u0026amp; Hamdi, H. (2024). The effect of mobile-assisted language learning (MALL) on grammar proficiency and learning motivation. \u003cem\u003eIslamic Manuscript of Linguistics and Humanity, 6\u003c/em\u003e(2), 23–33.\u003c/li\u003e\n \u003cli\u003eField, J. (2010). Listening in the language classroom. \u003cem\u003eELT Journal\u003c/em\u003e, 64(3), 331–333. https://doi.org/10.1093/elt/ccq018\u003c/li\u003e\n \u003cli\u003eGarcía Botero, G., Questier, F., \u0026amp; Zhu, C. (2019). Self-directed language learning in a mobile-assisted, out-of-class context: Do students walk the talk? \u003cem\u003eComputer Assisted Language Learning, 32\u003c/em\u003e(1–2), 71–97. https://doi.org/10.1080/09588221.2018.1485707\u003c/li\u003e\n \u003cli\u003eGarcía-Carrión, R., López de Aguileta, G., Padrós, M., \u0026amp; Ramis-Salas, M. (2020). Implications for social impact of dialogic teaching and learning. \u003cem\u003eFrontiers in Psychology\u003c/em\u003e, 11, 140. https://doi.org/10.3389/fpsyg.2020.00140\u003c/li\u003e\n \u003cli\u003eGarcía Botero, G., Questier, F., \u0026amp; Zhu, C. (2018). Self-directed language learning in a mobile-assisted, out-of-class context: do students walk the talk? \u003cem\u003eComputer Assisted Language Learning\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(1–2), 71–97. https://doi.org/10.1080/09588221.2018.1485707\u003c/li\u003e\n \u003cli\u003eGhaed Sharaf, F., Mobaraki, M., \u0026amp; Rabani Nia, M. (2018). The effect of listening strategy instruction on Iranian pre-intermediate EFL learners’ listening ability. \u003cem\u003eInternational Journal of Applied Linguistics and English Literature, 7\u003c/em\u003e(3), 188–192. https://doi.org/10.7575/aiac.ijalel.v.7n.3p.188\u003c/li\u003e\n \u003cli\u003eGhasemi, M. J., \u0026amp; Mirzapour Kouhdasht, A. (2024). Beyond the ratings: Exploring Duolingo's perceived usefulness among language learners. \u003cem\u003eTechnology Assisted Language Education, 2\u003c/em\u003e(1), 57–70. https://doi.org/10.22126/tale.2024.10689.1039\u003c/li\u003e\n \u003cli\u003eGoh, C. C., \u0026amp; Vandergrift, L. (2021). \u003cem\u003eTeaching and learning second language listening: Metacognition in action\u003c/em\u003e. Routledge.\u003c/li\u003e\n \u003cli\u003eGoodwin, S., \u0026amp; Naismith, B. (2023). Assessing listening on the Duolingo English Test. \u003cem\u003eDuolingo Research Report DRR-23-02\u003c/em\u003e. Duolingo. http://duolingo-testcenter.s3.amazonaws.com/media/resources/listening-whitepaper.pdf\u003c/li\u003e\n \u003cli\u003eGragera, R. (2024). Motivation and proficiency in EFL: A case study using Duolingo. \u003cem\u003eWorld Journal of Technology, Learning, and Education, 1\u003c/em\u003e(1). https://doi.org/10.33422/worldtle.v1i1.671\u003c/li\u003e\n \u003cli\u003eHidri, S. (2019). Static vs. dynamic assessment of students’ writing exams: A comparison of two assessment modes. \u003cem\u003eInternational Multilingual Research Journal\u003c/em\u003e, 13(4), 239–256. https://doi.org/10.1080/19313152.2019.1654899\u003c/li\u003e\n \u003cli\u003eHosseini, S. A., Rajabi, P., \u0026amp; Mahmoodi, K. (2021). Flipped classroom model and its impact on Iranian EFL learners' classroom anxiety and listening performance. \u003cem\u003eComputer-Assisted Language Learning Electronic Journal, 22\u003c/em\u003e(3), 1–16.\u003c/li\u003e\n \u003cli\u003eHsu, Y. H. (2024). An analysis of the effectiveness of listening activities in improving English learning/reading (L2 language-English). In \u003cem\u003eSHS Web of Conferences\u003c/em\u003e (Vol. 200, p. 02022). EDP Sciences. https://doi.org/10.1051/shsconf/202420002022\u003c/li\u003e\n \u003cli\u003eHuynh, D., Zuo, L., \u0026amp; Iida, H. (2016). Analyzing gamification of “Duolingo” with focus on its course structure. In \u003cem\u003eGames and Learning Alliance: 5th International Conference, GALA 2016, Utrecht, The Netherlands, December 5–7, 2016, Proceedings 5\u003c/em\u003e (pp. 268–277). Springer International Publishing. https://doi.org/10.1007/978-3-319-50182-6_25\u003c/li\u003e\n \u003cli\u003eInayah, N., Yusuf, Q., \u0026amp; Fibula, N. (2020). Exploring undergraduate students’ perception toward the use of Duolingo in learning English. \u003cem\u003eHumanities \u0026amp; Social Sciences Reviews\u003c/em\u003e, 8(3), 76–85. https://doi.org/10.18510/hssr.2020.839\u003c/li\u003e\n \u003cli\u003eJašková, J. (2014). \u003cem\u003eDuolingo as a new language-learning website and its contribution to e-learning education\u003c/em\u003e (Doctoral dissertation, Masaryk University, Faculty of Education). \u003c/li\u003e\n \u003cli\u003eJiang, X., Peters, R., Plonsky, L., \u0026amp; Pajak, B. (2024). The effectiveness of Duolingo English courses in developing reading and listening proficiency. \u003cem\u003eCALICO Journal\u003c/em\u003e, 41(3), 1–24. https://doi.org/10.1558/cj.26704\u003c/li\u003e\n \u003cli\u003eJiang, X., Peters, R., \u0026amp; Pajak, B. (2023). Reading and listening outcomes of learners in the Duolingo English course for Portuguese speakers (Vol. 1). \u003cem\u003eDuolingo Research Report DRR-23\u003c/em\u003e. https://duolingo-papers.s3.amazonaws.com/reports/duolingo-efficacy-english-reading-listening-whitepaper.pdf\u003c/li\u003e\n \u003cli\u003eJiang, X., \u0026amp; Pajak, B. (2022). Reading and listening outcomes of learners in the Duolingo English course for Spanish speakers (pp. 1–10). \u003cem\u003eDuolingo Research Report DRR-22-08\u003c/em\u003e. https://duolingo-papers.s3.amazonaws.com/reports/duolingo-english-efficacy-whitepaper.pdf\u003c/li\u003e\n \u003cli\u003eLanders, R. N. (2019). Gamification misunderstood: How badly executed and rhetorical gamification obscures its transformative potential. \u003cem\u003eJournal of Management Inquiry, 28\u003c/em\u003e(2), 137–140. https://doi.org/10.1177/1056492618790913\u003c/li\u003e\n \u003cli\u003eLantolf, J. P. (2000). Sociocultural theory and second language learning. \u003cem\u003eNew York Oxford University Press Google Scholar\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 29-63.\u003c/li\u003e\n \u003cli\u003eLantolf, J. P., \u0026amp; Poehner, M. E. (2014). \u003cem\u003eSociocultural theory and the pedagogical imperative in L2 education: Vygotskian praxis and the research/practice divide\u003c/em\u003e. Routledge.\u003c/li\u003e\n \u003cli\u003eLantolf, J. P., Poehner, M. E., \u0026amp; Thorne, S. L. (2020). Sociocultural theory and L2 development. In B. VanPatten, G. D. Keating, \u0026amp; S. Wulff (Eds.), \u003cem\u003eTheories in second language acquisition: An introduction\u003c/em\u003e (3rd ed., pp. 223–247). Routledge. https://doi.org/10.4324/9780429503986-10\u003c/li\u003e\n \u003cli\u003eLee, J. S. (2019). EFL students’ views of willingness to communicate in the extramural digital context. \u003cem\u003eComputer Assisted Language Learning\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(7), 692–712. https://doi.org/10.1080/09588221.2018.1535509\u003c/li\u003e\n \u003cli\u003eLi, Z., \u0026amp; Bonk, C. J. (2023). Self-directed language learning with Duolingo in an out-of-class context. \u003cem\u003eComputer Assisted Language Learning\u003c/em\u003e, 1–23. https://doi.org/10.1080/09588221.2023.2206874\u003c/li\u003e\n \u003cli\u003eMayer, R. E. (2024). The past, present, and future of the cognitive theory of multimedia learning. \u003cem\u003eEducational Psychology Review, 36\u003c/em\u003e(1), 8. https://doi.org/10.1007/s10648-023-09842-1\u003c/li\u003e\n \u003cli\u003eMa, X., \u0026amp; Zhang, H. (2024). Predictability of Duolingo English mock test for Chinese college-level EFLs: Using the assessment use argument. \u003cem\u003eFrontiers in Education\u003c/em\u003e, 8, 1275518. https://doi.org/10.3389/feduc.2023.1275518\u003c/li\u003e\n \u003cli\u003eNamaziandost, E., Neisi, L., Mahdavirad, F., \u0026amp; Nasri, M. (2019). The relationship between listening comprehension problems and strategy usage among advanced EFL learners. \u003cem\u003eCogent Psychology, 6\u003c/em\u003e(1), 1691338. https://doi.org/10.1080/23311908.2019.1691338\u003c/li\u003e\n \u003cli\u003eNushi, M., \u0026amp; Eqbali, M. H. (2017). Duolingo: A mobile application to assist second language learning. \u003cem\u003eTeaching English with Technology\u003c/em\u003e, 17(1), 89–98. \u003c/li\u003e\n \u003cli\u003eQub’a, A. A., Al-Haj Eid, O. A. A., Hasan, G. A., \u0026amp; Al Herz, J. (2024). The effect of utilizing gamification in enhancing English language skills in university settings. \u003cem\u003eWorld Journal of English Language, 14\u003c/em\u003e(4), 428–436. https://doi.org/10.5430/wjel.v14n4p428\u003c/li\u003e\n \u003cli\u003ePoehner, M. E. (2009). Dynamic assessment as a dialectic framework for classroom activity: Evidence from second language (L2) learners. \u003cem\u003eJournal of Cognitive Education and Psychology, 8\u003c/em\u003e(3), 252–268. https://doi.org/10.1891/1945-8959.8.3.252\u003c/li\u003e\n \u003cli\u003ePoehner, M. E., \u0026amp; van Compernolle, R. A. (2014). Frames of interaction in dynamic assessment: Developmental diagnoses of second language learning. In \u003cem\u003eAddressing issues of access and fairness in education through dynamic assessment\u003c/em\u003e (pp. 89–104). Routledge.\u003c/li\u003e\n \u003cli\u003ePurwanto, H. N., Faridi, A., \u0026amp; Rozi, F. (2022). The effect of Duolingo and SPADA to teach listening to students with different achievement levels. \u003cem\u003eEnglish Education Journal, 12\u003c/em\u003e(1), 87–95. https://doi.org/10.15294/eej.v12i1.54417\u003c/li\u003e\n \u003cli\u003ePutri, L. M., \u0026amp; Islamiati, A. (2018). Teaching listening using Duolingo application. \u003cem\u003eJournal of English Education, 1\u003c/em\u003e(4), 460. https://doi.org/10.31327/jee.v1i4.460\u003c/li\u003e\n \u003cli\u003eRost, M., \u0026amp; Brown, S. (2022). Second language listening. In \u003cem\u003eHandbook of practical second language teaching and learning\u003c/em\u003e (pp. 238–255). Routledge. https://doi.org/10.4324/9781003191314-17\u003c/li\u003e\n \u003cli\u003eSansone, C., Geerling, D. M., Thoman, D. B., \u0026amp; Smith, J. L. (2019). Self-regulation of motivation: A renewable resource for learning. In K. A. Renninger \u0026amp; S. Hidi (Eds.), \u003cem\u003eThe Cambridge Handbook of Motivation and Learning\u003c/em\u003e (pp. 87–110). Cambridge University Press. https://doi.org/10.1017/9781316823279.006\u003c/li\u003e\n \u003cli\u003eShortt, M., Tilak, S., Kuznetcova, I., Martens, B., \u0026amp; Akinkuolie, B. (2023). Gamification in mobile-assisted language learning: A systematic review of Duolingo literature from public release of 2012 to early 2020. \u003cem\u003eComputer Assisted Language Learning\u003c/em\u003e, 36(3), 517–554. https://doi.org/10.1080/09588221.2021.1933540\u003c/li\u003e\n \u003cli\u003eSwain, M., Kinnear, P., \u0026amp; Steinman, L. (2015). \u003cem\u003eSociocultural theory in second language education: An \u003c/em\u003e\u003c/li\u003e\n \u003cli\u003eSzabó, F., \u0026amp; Kopinska, M. (2023). Gamification in foreign language teaching: A conceptual introduction. \u003cem\u003eHungarian Educational Research Journal, 13\u003c/em\u003e(3), 418–428. https://doi.org/10.1556/063.2023.00202\u003c/li\u003e\n \u003cli\u003eTajik, A. (2024). Exploring the potential of ChatGPT in EFL language learning: Learners’ reflections and practices. \u003cem\u003ePreprints\u003c/em\u003e. https://doi.org/10.20944/preprints202412.2218.v1\u003c/li\u003e\n \u003cli\u003eTajik, A. (2025). Gamified and non-gamified AI tools in enhancing EFL listening comprehension: An analysis of Duolingo and Replika’s impact on engagement, motivation, and learning outcomes. https://doi.org/10.21203/rs.3.rs-6032009/v1\u003c/li\u003e\n \u003cli\u003eTajik, A. (2025). Exploring the role of AI-driven dynamic writing platforms in improving EFL learners' writing skills and fostering their motivation. https://doi.org/10.21203/rs.3.rs-5788599/v1\u003c/li\u003e\n \u003cli\u003eTajik, A. (2025). Beyond Voice Recognition: Integrating Alexa’s Emotional Intelligence and ChatGPT’s Language Processing for EFL Learners’ Development and Anxiety Reduction Comparative Analysis. https://doi.org/10.21203/rs.3.rs-5989702/v1\u003c/li\u003e\n \u003cli\u003eTorres Rodríguez, D. A., Armijos Ramírez, M. R., Criollo Vargas, M. I., \u0026amp; Salazar Chamba, E. M. (2023). Gamification strategies on the development of English listening comprehension skills. \u003cem\u003eRevista Multidisciplinaria Investigación Contemporánea, 1\u003c/em\u003e(2), 30–57. https://doi.org/10.58995/redlic.ic.v1.n2.a51\u003c/li\u003e\n \u003cli\u003eTuong, N. K., \u0026amp; Dan, T. C. (2024). A study on Duolingo mobile applications to improve EFL students' listening comprehension performances. \u003cem\u003eEuropean Journal of Alternative Education Studies, 9\u003c/em\u003e(1), 217–265. https://doi.org/10.46827/ejae.v9i1.5342\u003c/li\u003e\n \u003cli\u003eVan Compernolle, R. A., \u0026amp; Zhang, H. (2014). Dynamic assessment of elicited imitation: A case analysis of an advanced L2 English speaker. \u003cem\u003eLanguage Testing\u003c/em\u003e, 31(4), 395–412. https://doi.org/10.1177/0265532214530984\u003c/li\u003e\n \u003cli\u003eVan de Pol, J., Mercer, N., \u0026amp; Volman, M. (2019). Scaffolding student understanding in small-group work: Students' uptake of teacher support in subsequent small-group interaction. \u003cem\u003eJournal of the Learning Sciences\u003c/em\u003e, 28(2), 206–239. https://doi.org/10.1080/10508406.2018.1522258\u003c/li\u003e\n \u003cli\u003eVenkatesh, V., Morris, M. G., Davis, G. B., \u0026amp; Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. \u003cem\u003eMIS Quarterly, 27\u003c/em\u003e(3), 425–478. https://doi.org/10.2307/30036540\u003c/li\u003e\n \u003cli\u003eWood, J. (2021). A dialogic technology-mediated model of feedback uptake and literacy. \u003cem\u003eAssessment \u0026amp; Evaluation in Higher Education\u003c/em\u003e, 46(8), 1173–1190. https://doi.org/10.1080/02602938.2020.1852174\u003c/li\u003e\n \u003cli\u003eXia, L., Shen, W., Fan, W., \u0026amp; Wang, G. A. (2024). Knowledge-Aware Learning Framework Based on Schema Theory to Complement Large Learning Models. \u003cem\u003eJournal of Management Information Systems\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(2), 453–486. https://doi.org/10.1080/07421222.2024.2340827\u003c/li\u003e\n \u003cli\u003eYıldırım, S., \u0026amp; Yıldırım, Ö. (2016). The importance of listening in language learning and listening comprehension problems experienced by language learners: A literature review. \u003cem\u003eAbant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi, 16\u003c/em\u003e(4), 2094–2110.\u003c/li\u003e\n \u003cli\u003eZimmerman, B. J. (2002). Becoming a self-regulated learner: An overview. \u003cem\u003eTheory into Practice, 41\u003c/em\u003e(2), 64–70. https://doi.org/10.1207/s15430421tip4102_2\u003c/li\u003e\n \u003cli\u003eZhang, S., \u0026amp; Hasim, Z. (2023). Gamification in EFL/ESL instruction: A systematic review of empirical research. \u003cem\u003eFrontiers in Psychology, 13\u003c/em\u003e, 1030790. https://doi.org/10.3389/fpsyg.2022.1030790\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Gamified Smart Learning Systems, EFL Listening Comprehension, Personalized Learning Paths, Learning Analytics, Self-efficacy, Adaptive Algorithms","lastPublishedDoi":"10.21203/rs.3.rs-7558517/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7558517/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eListening comprehension is not only a vital pillar of English language acquisition but also one of the most underexplored skills in adolescent EFL education. This study offers a novel, head-to-head comparison of two AI-powered platforms with contrasting pedagogical designs\u0026mdash;\u003cb\u003eDuolingo\u003c/b\u003e, a gamified system built around rewards and progress tracking, and \u003cb\u003eReplika\u003c/b\u003e, a non-gamified, conversation-based AI. Addressing a clear gap in the literature, we examine how these divergent approaches influence high school learners\u0026rsquo; listening comprehension, engagement, and motivation. Using a concurrent mixed-methods design, we worked with 53 Iranian high school students (aged 14\u0026ndash;15), randomly assigned to a Duolingo group (n\u0026thinsp;=\u0026thinsp;27) or a Replika group (n\u0026thinsp;=\u0026thinsp;26) in a 12-week, 24-session program. Quantitative data came from researcher-developed pre- and post-tests and perception questionnaires; qualitative insights were drawn from classroom observations and semi-structured interviews (n\u0026thinsp;=\u0026thinsp;15). Results revealed that while both groups achieved significant gains in listening comprehension, Duolingo\u0026rsquo;s adaptive gamification consistently drove higher engagement and motivation. Replika, in contrast, nurtured more sustained, reflective listening through personalized dialogue. These findings show that design architecture\u0026mdash;whether gamified or conversational\u0026mdash;matters for how AI can address long-standing gaps in EFL instruction. Beyond improving listening comprehension, the study illustrates how targeted integration of AI-driven strategies can help educators, curriculum designers, and EdTech developers create more engaging, context-sensitive, and sustainable language learning experiences.\u003c/p\u003e","manuscriptTitle":"AI-Driven Gamified and Non-Gamified Platforms in EFL Listening: Comparative Effects of Duolingo and Replika on Comprehension, Engagement, and Motivation","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-09 14:46:46","doi":"10.21203/rs.3.rs-7558517/v1","editorialEvents":[{"type":"communityComments","content":18}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"8278a919-cff6-4e48-ac6f-5d9f55e0cd0c","owner":[],"postedDate":"September 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-09-18T07:53:47+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-09 14:46:46","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7558517","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7558517","identity":"rs-7558517","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}