REAL-TIME AI FEEDBACK VS. TEACHER FEEDBACK ON LISTENING TASKS: A COMPARATIVE STUDY

Аннотация

This study explores the comparative impact of real-time Artificial Intelligence (AI) feedback and traditional teacher feedback on learners’ performance in English as a Foreign Language (EFL) listening tasks. As AI technologies gain ground in educational environments, their capacity to offer immediate, data-driven, and personalized feedback has positioned them as potential complements or even alternatives to human instruction. The research draws on a six-week intervention with Uzbek EFL learners, comparing two groups: one using AI-integrated listening platforms and the other receiving direct teacher feedback. Through quantitative assessments and qualitative observations, the study investigates comprehension gains, learner engagement, and affective responses. Findings reveal that AI feedback supports faster self-correction and increases student autonomy, while teacher feedback fosters deeper comprehension and emotional connection. Both feedback types contribute positively to learning, albeit in different ways. The paper concludes that a blended approach—integrating AI precision with human empathy—offers the most pedagogically effective solution for developing listening skills in EFL contexts. Implications for classroom practice, teacher training, and AI tooldesign are discussed. Keywords : Artificial Intelligence, real-time feedback, teacher feedback, listening comprehension, EFL learning, educational technology, learner engagement, formative assessment, feedback effectiveness, Uzbekistan. As language education continues to evolve in the digital era, the integration of Artificial Intelligence (AI) into classroom practices is reshaping how students engage with language skills, especially listening. One of the most discussed innovations is the use of real-time AI-generated feedback during listening tasks. Unlike traditional teacher feedback, which is typically delayed and context- bound, AI systems can instantly evaluate learners’ responses, offer corrections, and provide performance analytics. This comparative study examines the effectiveness of real-time AI feedback versus teacher-provided feedback in the context of English as a Foreign Language (EFL) listening instruction. Drawing on classroom experiments and learner reflections, the study explores the pedagogical implications, learner preferences, and outcome differences between the two modes of feedback. AI-based feedback systems utilize speech recognition and natural language processing to evaluate pronunciation, comprehension accuracy, and task completion. In listening tasks, these systems can track how many times a student listens to an audio, where they pause or replay, and how accurately they respond to questions. Immediate feedback is then generated in the form of scores, suggestions, or prompts for improvement. This immediacy encourages students to reflect on their performance while the task is still cognitively active, promoting self-correction and deeper learning. Moreover, AI tools are consistent, tireless, and available outside class time, giving students autonomy and flexibility inpractice. In contrast, teacher feedback carries the advantage of human nuance, empathy, and pedagogical intuition. Teachers can tailor their comments to individual learning styles, emotional states, and contextual factors. They are able to clarify misunderstandings, use simplified explanations, and build a supportive learning atmosphere. Moreover, teacher feedback often incorporates motivational elements and a relational dimension, which AI lacks. However, teacher feedback is generally worldlyjournals.com ILMIY TADQIQOTLAR VA ULARNING YECHIMLARI JURNALI JOURNAL OF SCIENTIFIC RESEARCH AND THEIR SOLUTIONS VOLUME 6, ISSUE 01, IYUL 2025 WORLDLY KNOWLEDGE NASHRIYOTI delayed, limited by classroom time constraints, and varies in consistency depending on the teacher’s workload and observation skills. Data from a sample of 100 Uzbek EFL learners aged 13 to 17 was analyzed over a six-week period. One group used AI-integrated listening apps that provided real-time feedback, while the other received traditional feedback from their teachers during regular classroom instruction. Pre- andpost- listening comprehension tests were administered, alongsidelearner surveys and teacherobservations. Results indicate that students in the AI feedback group showed faster improvements in self- correction behaviors and developed greater awareness of listening strategies. They were more likely to reattempt tasks and adjust their listening habits based on automated hints. Many reported enjoying the independence and immediacy provided by AI tools, which motivated them to practice more frequently. However, they also expressed occasional frustration when AI feedback was overly mechanical, misinterpreted theirinput, orfailed toexplain errors clearly. Teacher feedback group showed more gradual progress, but with deeper comprehension and better integration of listening with speaking and grammar. Students