Dirbtinio intelekto naudojimo Lietuvos gimnazijose mąstai, rizikos ir galimybės
Recenzentas / Rewiewer |
Licencinė sutartis Nr. MRU-EDT-1900.
-.
Topic: Scales, risks, and opportunities of AI use in Lithuanian gymnasiums. Aim: to examine the extent of artificial intelligence (AI) use in Lithuanian gymnasiums and to analyse the challenges encountered while integrating AI into the educational process. Methodology. The empirical research was conducted using a quantitative survey method. An online questionnaire was administered to 399 teachers representing different school locations, age groups, teaching experience levels, and subject areas. The instrument included Likert-scale items measuring AI usage intensity, purposes of application, perceived benefits, and challenges. Descriptive and inferential statistical methods were applied, including mean comparisons and reliability testing using Cronbach’s alpha (ranging from α = 0.89 to α = 0.92), confirming high internal consistency of the scales. The study also tested three hypotheses: H1 – Teachers in Lithuanian gymnasiums actively integrate AI tools into the educational process; H2 – Teachers’ age is significantly associated with the intensity of AI use; H3 – Teachers with longer professional experience use AI tools less frequently. Results. The findings reveal that AI integration in Lithuanian gymnasiums remains fragmented. Teachers most often use AI for information search and idea generation, less frequently for lesson planning, and least for student self-assessment, feedback, gamification, or personalisation. AI use is predominantly episodic rather than systematic. The ecosystem of tools is narrow: teachers overwhelmingly rely on ChatGPT, while specialised AI solutions, such as Gemini, MagicSchool AI, Kahoot/Quizizz with AI features, Mentimeter, or adaptive learning platforms, are rarely used. Although teachers acknowledge the benefits of AI for reducing workload and improving lesson preparation efficiency, they remain cautious regarding its effect on students’ motivation and learning outcomes. Barriers include a lack of training, insufficient methodological guidance, inadequate school infrastructure, and concerns about academic integrity and data protection. Hypothesis testing showed that H1 was not confirmed, while H2 and H3 were supported: younger teachers and those with shorter teaching experience use AI more frequently and more diversely, whereas older and long-serving teachers adopt AI significantly less often.