Investigating the Impact of Integrating Prompting Strategies in ChatGPT on Students Learning Achievement and Cognitive Load

This study primarily investigates integrating prompting strategies into ChatGPT in instructional applications in elementary school natural science curricula. It examines the effects of different learning strategies (prompting strategies integrated into the ChatGPT natural science instructional group and worksheet tasks integrated into the ChatGPT natural science instructional group) on elementary students’ learning achievement and cognitive load in natural science. This study adopted a quasi-experimental research design involving fifty fifth-grade students from two classes in a primary school in Taichung City. The classes are divided into two groups: one integrates prompting strategies into ChatGPT natural science instructional, and the other integrates worksheet tasks into ChatGPT natural science instructional. Both groups use ChatGPT as a learning aid within three sessions of instructional experiments. Pre- and post-tests of natural science learning achievement are conducted before and after the experiments. Additionally, the study explores the impact of integrating prompting strategies into ChatGPT instruction on elementary students’ cognitive load in natural science. The results of this study reveal that: (1) integrating prompting strategies into ChatGPT instruction facilitates the enhancement of elementary students’ natural science learning achievement, and (2) the use of a guided questioning system in prompting strategies integrated into ChatGPT natural science instruction demonstrates the potential to reduce students’ cognitive load.