The effectiveness of ChatGPT on pre-service teachers' STEM teaching literacy, learning performance, and cognitive load in a teacher training course

Abstract

Within the realm of teacher education, effective tools as support can enhance pre-service teachers' comprehension of STEM concepts and their ability to tackle with complex problems. This research aims to empirically investigate the impact of ChatGPT-supported collaborative learning on pre-service teachers' STEM teaching literacy, learning performance, and cognitive load. A total of 36 sophomore undergraduate pre-service teachers participated in an 8-week quasi-experimental study (with classes lasting 3 hours each week). These pre-service teachers were tasked with creating teaching aids and simulating instructional sessions. The experimental group engaged in small-group collaborative learning assisted by ChatGPT (N=21), while the control group (N=15) conducted conventional collaborative learning. The results indicate that pre-service teachers using ChatGPT demonstrated superior performance in terms of learning performance and cognitive load, though the improvement in STEM teaching literacy was not evident. However, ChatGPT contributed to students' acquisition and comprehension of STEM teaching knowledge. In essence, within a collaborative learning context, ChatGPT exhibits significant potential in addressing complex issues and offering personalized support but also we should mitigate potential risks. Collaborative learning with human-machine co-creation is poised to be a trend in the future of education, offering innovative avenues for the cultivation of STEM pre-service teachers.