Investigating the Effects of Prompt Engineering in STEM Activities Designed with ChatGPT-4o

Abstract

This study examines the use of ChatGPT-4o, an artificial intelligence-supported chatbot, in the design of STEM activities and investigates the effects of prompt engineering practices on this process. The aim of the study is to determine how pre-service mathematics teachers with no prior experience in designing STEM activities navigate ChatGPT-4o, how the activities they design fit into the STEM framework, and how prompt engineering techniques influence the quality of the activities. The research was conducted with the case study method with seven pre-service mathematics teachers enrolled in an elementary mathematics teacher education program at a university in Turkey. During the data collection process, the pre-service teachers designed STEM activities using ChatGPT-4o, first without knowledge of prompt engineering. Following a five-hour instructional session, they repeated the task using prompt engineering strategies. The designed activities were analyzed using the “STEM Activity Evaluation Form.” The findings showed that STEM activities designed without knowledge of prompt engineering were generally superficial in terms of thematic depth. However, after receiving the prompt engineering training, pre-service teachers produced more appropriate and interdisciplinary STEM activities. The results of the study reveal that ChatGPT-4o has strong potential as a guide in STEM education, but prompt engineering is critical for its effective use. The researchers recommend the development of prompt engineering skills into teacher education and the effective integration of artificial intelligence tools in STEM instruction.

Ähnliche Arbeiten