Chatgpt Support Vs. Dependence: Striking the Balance in Infuencing the Learning Outcomes of Beed Students

Abstract

This study explored second year BEEd students' perspective of ChatGPT as an assistive educational tool and as an enabling source of dependence. The researcher used a quantitative descriptive-comparative design to understand some of the opportunities and limitations of ChatGPT, as perceived by the students, around their academic learning and performance. Data were collected using a survey questionnaire and by a documentary analysis of GWA between two groups of students; who used ChatGPT as a learning aid and for who used it as an adjunct to their learning. Survey results showed agreement on writing, grammar, and conceptual understanding (M = 3.79–3.89) but neutral concerns about misinformation and motivation. Supplementary users averaged a 90.93 GWA (Very Satisfactory) while dependent users averaged 88.84 GWA (Satisfactory); the difference was significant (t = 2.36, p = 0.0215). Overall mean GWA = 89.27 ± 2.48 fell short of the mastery benchmark (90). The study concludes that ChatGPT enhances learning only when used as a scaffold, not as a crutch, and recommends monitored, guided integration. The study briefly theorizes the balancing of AI integration by using Technology Acceptance Model (TAM), Student Development Theory (SDT), and the framework of Vygotsky's ZPD, and recommend educators and students professionally moderate the use of ChatGPT in education to utilize it responsibly so that students learn with technology, without compromising academic integrity and cognitive development. This paper concludes with a comprehensive account of educational implications for educators, policymakers, and learners for the advancement of AI in their learning and education.