20.500.12556/DKUM-95338 Slovenski srednješolski učitelji jezikov in veliki jezikovni modeli magistrsko delo Slovenian secondary school language teachers and large language models Pričujoče delo obravnava uporabo velikih jezikovnih modelov (LLM) pri pouku tujih jezikov na slovenskih srednjih šolah z vidika učiteljev. Osrednji poudarek je na vplivu LLM na pedagoško delo, doseganju učnih ciljev ter prilagajanju učnega procesa učiteljev ob zavedanju, da dijaki ta orodja uporabljajo pri domačem delu. Raziskava je bila izvedena z anketnim vprašalnikom med 64 učitelji jezikov. Rezultati kažejo, da 77 % učiteljev pri svojem delu uporablja LLM, najpogosteje ChatGPT. Učitelji jih največ uporabljajo za pripravo gradiv in diferenciacijo pouka, redkeje pa za ocenjevanje, povratne informacije in administrativne naloge. Večina zaznava, da dijaki LLM uporabljajo pogosto ali občasno, zato se na to odzivajo s prilagajanjem nalog, večjim poudarkom na procesih učenja ter pogovori o koristih in pasteh umetne inteligence. Statistično značilnih razlik med učitelji glede na delovno dobo, program ali predmet ni bilo zaznati, potrjena pa je bila pozitivna povezava med pogostostjo uporabe in občutkom kompetentnosti. Ugotovitve nakazujejo, da LLM prinašajo učiteljem pomembne priložnosti, a hkrati zahtevajo premišljene pedagoške prilagoditve in smernice. This thesis examines the use of large language models (LLM) in foreign language teaching at Slovenian secondary schools from the perspective of teachers. The focus is on how LLMs affect pedagogical work, the achievement of learning objectives, and how teachers adapt the learning process with the awareness that students make use of these tools when completing homework. The study was carried out through a questionnaire survey among 64 language teachers. Results show that 77% of teachers employ LLMs in their work, most frequently ChatGPT. LLMs are used mainly for lesson preparation and instructional differentiation, while they are less frequently applied in assessment, feedback, or administrative tasks. Teachers largely perceive that students use LLMs frequently or occasionally, and respond by adjusting assignments, emphasizing learning processes, and discussing the benefits and risks of artificial intelligence. No statistically significant differences were found between teachers regarding years of experience, programme, or subject taught. However, a positive correlation was confirmed between frequency of use and perceived competence, indicating that greater use of LLMs enhances teachers’ sense of digital confidence. The findings suggest that LLMs provide important opportunities for teachers but also demand thoughtful pedagogical adaptations and clear guidelines. veliki jezikovni modeli umetna inteligenca generativna umetna inteligenca učenje jezikov srednješolsko izobraževanje large language models artificial intelligence generative artificial intelligence language learning secondary education true true false U. Kajbič Slovenski jezik Angleški jezik Magistrsko delo/naloga 2025-09-14 23:41:46 2025-09-29 14:38:02 2025-09-30 03:34:52 0000-00-00 00:00:00 2025 0 Maribor 0 Maribor 1 spletni vir (1 datoteka PDF (X, 73 str.)) 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 37.091.3:004.8(043.2)(0.034.2) 250921731 MAG_Kajbic_Urska_2025.pdf MAG_Kajbic_Urska_2025.pdf 1 FF9C65BF0F641A4BCC79E52A290182DE 1689cf9fc3c719dc0fdb28304719b6af448a2e27e605d26b9592a525535307bc 4f115bca-91b5-11f0-b621-00155d000105 https://dk.um.si/Dokument.php?lang=slv&id=198391 Filozofska fakulteta 0 0 0