ÇEVRİMİÇİ ÖĞRENMEDE ÜRETKEN YAPAY ZEKÂ DESTEKLİ SOHBET ROBOTUNUN TASARLANMASI VE DEĞERLENDİRİLMESİ

Abstract

Online learning has become one of the most widely adopted approaches in contemporary education, though it presents challenges such as the transactional distance experienced by learners. Generative AI-powered chatbots, with their capacity for instant feedback and personalized support, have been regarded as potential tools for reducing this distance. This study aimed to examine the effect of a generative AI-powered chatbot on transactional distance in online learning and the dynamics shaping this effect. The study was conducted in two phases: a chatbot was first designed and developed through design-based research, then evaluated using an explanatory sequential mixed-methods design examining usage data, perceptions of transactional distance, and student experiences. Results showed that transactional distance was lower in learner-interface, learner-content, learner-instructor, and learner-environment interactions in the online learning environment where the chatbot was present, with no significant difference in learner-learner interaction. Message intensity emerged as the only consistent predictor of transactional distance, specifically in learner-instructor and learner-environment interactions. Qualitative findings revealed that, depending on usage intensity, students differed in how they positioned the chatbot in the learning process, the trust they placed in it, and the social meaning they attributed to it. Overall, the findings suggest that generative AI-powered chatbots can reduce transactional distance while reconfiguring the theory's dialogue and autonomy components; accordingly, the study proposes an extension to Transactional Distance Theory through the concepts of learner-driven dialogue and operational and cognitive autonomy.